Local hunters known as Vigilantes armed with locally made guns are seen on a pickup truck in Yola city of Adamawa State in Nigeria, on 6 December 2014. Mohammed Elshamy/ANADOLU AGENCY
Report 251 / Africa 20+ minutes

Double-edged Sword: Vigilantes in African Counter-insurgencies

Vigilante groups have been successful in providing local security. But subcontracting security functions to vigilante groups for counter-insurgency purposes is a dangerous option for fragile African states. African leaders should set clear objectives and mandates when enlisting vigilantes and invest in disarmament, demobilisation and reintegration programs.

  • Share
  • Save
  • Print
  • Download PDF Full Report

Executive Summary

As weak African states face growing insurgencies, they do what weak states tend to do: subcontract certain security functions to non-state actors or vigilante groups, many of which had taken up arms to protect their communities. This approach at times is viewed as a necessity, but is often dangerous, particularly in politically fluid and fractious states. The more fragile the state, the more it is dependent on vigilantes, but also the less able it is to police them or prevent abuse of power. The more successful the vigilante group against insurgents, the harder it is to demobilise, and the more likely it will become entrenched. As a result of ethnic rivalries and allegiances, community defence groups can morph into predatory, quasi-criminal organisations or enemies of the central state. Yet even when risks outweigh benefits, African leaders may not have the luxury of choice. At a minimum, African governments and their international backers should learn from the past, try to prevent abuses, guard against vigilantes’ mission creep and plan how to manage them once the conflict dies down.

By their very nature, vigilante groups carry inherent risk. Typically recruited from local communities, their members likely share the same ethnic or political identity, collective interests and threat perceptions, raising the odds that they will act as local militias – potentially more powerful than state authorities – and pursue narrow ethnic agendas; a short-term necessary evil that could pave the way for longer-term conflict. A solution for states in dire need of backing, vigilantes too often take advantage of their newfound capacity – and compensate for inadequate support and resources – by seeking to maximise their power and wealth through extortion, kidnapping, and other violent abuses.

But there are positive lessons to be learned too. Vigilante groups can be far more effective than state actors in providing local security. They generally enjoy greater legitimacy by virtue of community roots, and can be more efficient in identifying, tracking and combating insurgents thanks to their familiarity with local languages, geography and culture. Successfully managed by state authorities – and international actors – they can enable national leaders to forge lasting political pacts with provincial elites and bolster state legitimacy among local communities. In short, and while African and international policymakers rightfully may be concerned that empowering non-state forces will undermine the state, vigilantes also can serve as valuable intermediaries between local communities and central authorities.

Drawing on four illustrative cases – Sierra Leone, Uganda’s Teso region, South Sudan’s former Western Equatoria State and Nigeria’s north east – this report seeks to shed light on factors that determine vigilantes’ evolution and impact on security and stability with the objective of helping governments and their international partners navigate this dilemma.

Investing in sufficiently generous demobilisation and reintegration programs is key to offering vigilante members viable alternative livelihoods and due recognition.

Among these factors: regime neglect of, or hostility toward such groups (as in South Sudan) can give rise to new rebels, while unbridled state support (as in Sierra Leone) can empower armed groups controlled by strongmen and motivated in part by narrow self-interest. The clearer vigilantes’ objectives and mandate are set in advance, and the greater the oversight by national and local leaders, the state military and local communities, the more effective the group can be and the less likely it will veer away from community defence and counter-insurgency goals. This is more likely to occur in instances where the political interests of the central state and local leaders are roughly aligned (as in Uganda). By contrast, a less defined mandate – one that allows vigilantes to step into local governance roles – can be a recipe for trouble, prolonging the existence of vigilante groups and enlarging their scope, enabling them to consolidate their power and creating greater economic incentives for them to hold on to it. In the longer term, investing in sufficiently generous demobilisation and reintegration programs is key to offering vigilante members viable alternative livelihoods and due recognition. Transitioning selected members to community policing units also could help prevent their reactivation in more hostile guises.

Several broad lessons, each to be applied with due care for local conditions, emerge from the case studies. In particular, African leaders that enlist vigilante groups for counter-insurgency purposes should:

  • Engage local leaders with influence over vigilantes with the aim of settling on finite, mutually acceptable objectives within an overarching counter-insurgency strategy, and ensuring they provide political oversight over rank-and-file members;
  • Be clear upfront with vigilante leaders and foot soldiers as to what they should expect as reward for their efforts and compensation for any losses;
  • Provide vigilantes with adequate political and material support, including weapons when necessary, with the goal of ensuring they are able to pursue their objectives, thereby reducing the risk of extortion of resources from civilians;
  • Where possible, provide military oversight of, and ensure accountability for vigilantes’ abusive actions;
  • Put in place upfront a gender-sensitive plan to demobilise vigilantes once the insurgent threat has receded and to help them find work in locally-relevant sectors.

International donors and partners face a similar conundrum. They too should benefit from relatively strong state authorities enjoying a monopoly over the use of violence. But when the state is too weak to confront an insurgency alone, or when the insurgent group doubles up as a terrorist organisation threatening outside interests, the temptation will be great for international actors to support a militia or vigilante group – with or at times without the state’s consent. Those international actors’ interests would be best served by working in concert with state authorities, helping them manage relations with vigilante groups, cautioning against the pitfalls of unfettered support or counterproductive repression. To the extent international players interact with vigilante groups, they should avoid providing direct support, lest they weaken national authorities’ bargaining position. Instead, they should be willing to assist states with resources to better control vigilantes and more effectively demobilise and reintegrate them.

Reliance on vigilante groups often is a faute de mieux solution for states facing a threat they cannot address alone. But as the cases in this report illustrate, there are better and worse ways of doing so, and of ensuring that a short-term expedient not turn into a long-term headache.

Nairobi/Dakar/Brussels, 7 September 2017


I. Introduction

African states confronting insurgent groups face a dilemma when civilians mobilise and take up arms to protect their local communities. These forces can play a major role in fending off attacks and provide regular armed forces with critical local knowledge, thereby bolstering the effectiveness of counter-insurgency campaigns. But vigilante groups also can undermine central authority, widen conflict by targeting ethnic or political rivals or threaten longer-term stability by continuing as an autonomous armed force after the original conflict has subsided. To use them is to wield a double-edged sword.

This report examines four cases in sub-Saharan Africa: the Kamajors, who fought in Sierra Leone’s civil war (1991-2002); the Arrow Boys of Teso, who confronted the Lord’s Resistance Army (LRA) in eastern Uganda (2003-2007); the Zande Arrow Boys, who battled the LRA and later rebelled against South Sudan’s Dinka-led regime (2005-present); and the Civilian Joint Task Force, which has worked closely with the armed forces and police to counter Boko Haram in north-eastern Nigeria (2013-present).

Although primarily based on field research conducted in 2016 and early 2017 in Sierra Leone, Uganda, South Sudan and Nigeria, the report also incorporates analysis from Crisis Group’s past work, putting into wider geographic and historical perspective more than fifteen years of analysis regarding the conflict in Sierra Leone, the LRA in Uganda and subsequently the broader region, and Boko Haram in the Lake Chad basin.[fn]See Crisis Group Africa Report N°244, Watchmen of Lake Chad: Vigilante Groups Fighting Boko Haram, 23 February 2017.Hide Footnote This research also draws on Crisis Group’s wider research into curbing violent religious radicalism.[fn]See Crisis Group Special Report N°1, Exploiting Disorder: al-Qaeda and the Islamic State, 14 March 2016.Hide Footnote

Crisis Group uses the term vigilantes to refer to members of civilian self-defence groups, community defence forces and civil militias, which are formed to protect their communities from non-state or state actors or to combat insurgents. This term, widely used in the African context, is not meant to imply that their activities are illegal, even though they initially might have lacked state authorisation.

II. A Recurrent Policy Dilemma

Vigilante groups have formed and continue to exist in weak African states where governments are unable or unwilling to protect civilians from security threats ranging from large-scale insurgency, to political or ethnic violence, to low-level banditry.[fn]In Nigeria, the Bakassi Boys and O’odua People’s Congress are two notable examples of vigilante groups that have taken on policing-type roles and committed serious abuses. See “The Bakassi Boys: The Legitimization of Murder and Torture”, Human Rights Watch, May 2002, and “The O’odua People’s Congress: Fighting Violence with Violence”, Human Rights Watch, February 2003. In west-central Tanzania, the Sungusungu mobilised at village level in the face of increasingly violent cattle theft. See Michael L. Fleisher, “Sungusungu: State-Sponsored Village Vigilante Groups among the Kuria of Tanzania”, Africa: Journal of the International African Institute, vol. 70, no. 2 (2000), pp. 209-228. In Burkina Faso, where state authorities have proved unable to curb crime, Koglweogo groups have sought to uphold their own kind of justice and their growing influence has provoked heated debate on how the government should respond. See “Au Burkina Faso, les milices d’autodéfense échappent peu à peu au contrôle des autorités”, Le Monde, 29 May 2017, and “Burkina: Koglweogo, les justiciers de la brousse”, Jeune Afrique, 6 April 2016.Hide Footnote The nature of the threat shapes the kinds of activities that vigilantes undertake, whether counter-insurgency roles typically played by the military or more policing-type duties. Yet, regardless of circumstance, the phenomenon of vigilantes faces an essential problem: states too weak to provide security on their own are most prone to enlist non-state armed actors and delegate some local security functions to them, but also most likely to lack resources and capacity to control vigilantes and prevent them from abusing power for their own individual or group interest.

This report examines cases of vigilante groups formed in response to insurgent threats as opposed to general lawlessness, since vigilantism in non-conflict settings presents a related but different set of challenges and policy implications. The four cases – historical and current cases from West and East Africa – were selected to assess what factors ultimately determine the outcomes – positive or negative – of reliance on vigilantes. While Crisis Group does not claim that these form a representative sample of vigilantism in African conflicts, they cover a range of experiences, from the relatively positive (Arrow Boys in Teso, Uganda) to decisively harmful in terms of human suffering and political instability (Kamajors in Sierra Leone and Arrow Boys in South Sudan). Case selection also was informed by Crisis Group’s institutional expertise and fresh field research.

A. Kamajors in Sierra Leone

Over eleven years (1991-2002), one of Africa’s most brutal civil wars unfolded in Sierra Leone, killing tens of thousands and displacing up to a quarter of the population.[fn]“The Armed Conflict in Sierra Leone”, Human Rights Watch, 11 April 2012.Hide Footnote Among the most powerful fighting groups were the Kamajors, who evolved from bands of young men defending their villages to the core of a state-armed national militia fighting alongside both the regular army and foreign forces. The Kamajors (whose name means hunter in Mende, the predominant language and tribe in the Southern and Eastern provinces) became a highly divisive entity. Many Sierra Leoneans still revere them for their bravery in defending first their home areas and later a democratically-elected government.[fn]See Ned Dalby, “In Search of the Kamajors, Sierra Leone’s Civilian Counter-insurgents”, Crisis Group Commentary, 7 March 2017.Hide Footnote But they also are reviled as a brutal tribal militia, which looted and killed suspected rebel collaborators and further destabilised the country. Such diverse, but not necessarily incompatible, views reflect ethnic and political prejudice and how people’s experiences of the Kamajors differed over time and in different places.

The Kamajors’ trajectory over the course of the long war demonstrates how vigilante groups can be effective community protectors and, at times, military auxiliaries, principally by virtue of their superior local knowledge. It also illustrates the dangers of helping vigilantes become militarily powerful forces operating outside their communities, without adequate state monitoring or control, particularly in countries riven by ethno-political tensions.

1. From community protectors to unwieldy paramilitary force

Sierra Leone’s civil war began in the early 1990s as a battle between government forces and the Revolutionary United Front (RUF).[fn]The RUF was led by former army Corporal Foday Sankoh and operated out of Liberia, with the support of Charles Taylor, then a rebel leader and later Liberian president. For Crisis Group reporting during the conflict see Crisis Group Africa Reports N°28, Sierra Leone: Time for a New Military and Political Strategy, 11 April 2001; N°35, Sierra Leone: Managing Uncertainty, 24 October 2001; and Africa Briefing N°6, Sierra Leone: Ripe for Elections?, 19 December 2001. See also, Lansana Gberie, A Dirty War in West Africa: The RUF and the Destruction of Sierra Leone (Bloomington, 2005); and Paul Richards, Fighting for the Rain Forrest: War, Youth and Resources in Sierra Leone (Oxford, 1996).Hide Footnote Originally based in Liberia, the RUF launched attacks on both military and civilian targets, principally in the Eastern and Southern provinces. In response, local leaders started mobilising young men, including Kamajor hunters, to defend their home areas.[fn]See map of Sierra Leone in Appendix A. Doctor Alpha Lavalie, a professor, and Doctor Joe Demby, a medical doctor who would later become vice president, mobilised a group in Kenema in the Eastern province, while Allieu Kondewa, a medicinal healer, formed a vigilante group in Bonthe, Southern province. Crisis Group interviews, former adviser of Kamajor leader Sam Hinga Norman, Bo, 12 January 2017; former Vice President Joe Demby, Bo, 12 January 2017.Hide Footnote A former army captain and local chief, Sam Hinga Norman, organised youth around Bo, the country’s second largest city.[fn]Norman was the Jiama-Bongor chiefdom regent chief. The government had appointed him after the former chief died and until a new one could be elected. Crisis Group interview, Norman’s former adviser, Bo, 12 January 2017.Hide Footnote Thanks to his military experience and strength of character, Norman soon became the Kamajors’ national leader and figurehead. As fighting spread, other tribes formed defence groups in their areas, but the Kamajors in the south and east remained by far the largest and earned a reputation as the fiercest.[fn]The Donso community defence force formed among the Kono tribe in Kono district, Eastern province. The Gbethi and Kapra groups emerged in the Temne-dominated areas in the centre and west. The Tamaboro of the Kuranko, Limba and Yalunka tribes formed in Koinadugu district in the far north. Crisis Group interviews, Freetown and Bo, January 2017. See also Sierra Leone Truth and Reconciliation Commission Report, Volume 3A, Chapter 4: Nature of the Conflict, p. 542.Hide Footnote

The beleaguered government, recognising the local forces’ effectiveness and the usefulness of their local knowledge, allowed them to act as army auxiliaries, serving principally as guides and informants. But distrust between Kamajors and soldiers soon undermined cooperation. To counter the insurgency, the government rapidly expanded the army, quadrupling its numbers from about 3,000 before the war to approximately 13,000 by 1992.[fn]The exact figure is uncertain, given the number of “ghost” soldiers who collected pay or rations without serving. See Crisis Group Report, Sierra Leone: Time for a New Military and Political Strategy, op. cit., p. 6.Hide Footnote Rapid expansion, coupled with deficient leadership, training and equipment, saw some front-line troops become so-called sobels (soldier/rebels) who preyed on civilians, sometimes in collaboration with insurgents.[fn]Crisis Group Report, Sierra Leone: Time for a New Military and Political Strategy, op. cit., p. 6. Also, Crisis Group interviews, former Kamajors, Bo and Gerihun, January 2017.Hide Footnote In response, the Kamajors defended their communities against both rebels and soldiers.

To compensate for its military weakness, the government hired a private South African military company – Executive Outcomes – which fought rebels from 1995 to early 1997. They relied heavily on the Kamajors’ local expertise. Their joint operations ushered in a period of sufficient stability to allow elections to be held in February 1996; these brought the Mende-dominated Sierra Leone People’s Party (SLPP) to power. Norman, the Kamajors’ best-known leader, became deputy defence minister, and the state ramped up its support to the local defence forces. In 1996, to reassure those who believed the Kamajors were becoming the ruling party’s army, the government established a national umbrella organisation for all vigilante groups, known as the Civil Defence Forces (CDF). A central coordinating committee, including representatives from diverse tribal defence groups, used government funds to buy arms, ammunition, food and medical supplies, which it distributed to field units.[fn]Crisis Group interview, former National Security Adviser, Freetown, 11 January 2017.Hide Footnote

Despite this façade of national unity, the Civil Defence Forces’ ethnically distinct units operated largely independently of each other.[fn]Sierra Leone Truth and Reconciliation Commission Report, Volume 3A, Chapter 4: Nature of the Conflict, pp. 541-542.Hide Footnote The Kamajors remained numerically dominant, partly because Mendeland saw the most insurgent activity, and received the lion’s share of government resources. Jealousy and fear of these irregular, largely Mende, forces helped fuel further army discontent, prompting a May 1997 coup by junior soldiers who established the Armed Forces Revolutionary Council (AFRC) and invited the rebel RUF to join their government. “The SLPP tribal hunter militia, the Kamajors, received logistics and supplies far beyond their immediate needs”, wrote a coup leader, arguing that the ruling party was favouring a “private army over our armed forces”.[fn]Statement by Johnny Paul Koroma in August 1997 cited in Crisis Group Report, Sierra Leone: Time for a New Military and Political Strategy, op. cit., p. 7.Hide Footnote Only a small portion of the military remained loyal to the toppled government, now exiled in Conakry, Guinea. That government appealed to the Economic Community of West African States (ECOWAS) for help and regional troops deployed under the banner of the ECOWAS Monitoring Group (ECOMOG) pushed the rebels and Armed Forces Revolutionary Council out of Freetown in February 1998. While this made possible the government’s return, the war nonetheless dragged on for another four years.

According to a former British high commissioner, the Kamajor-dominated Civil Defence Forces were crucial to restoring state control.[fn]Crisis Group telephone interview, Peter Penfold, 16 January 2017.Hide Footnote It fought on behalf of the elected government, both independently and in coordination with ECOMOG troops. The Kamajors were a significant battlefield force in part due to their size and spread. The number of enrolees mushroomed to some 37,000 members, most of them rural, uneducated youth.[fn]The total number of CDF members at its height is hard to pin down given the group’s fluidity and informal recruitment, but at the end of the conflict over 37,000 individuals identifying themselves as CDF passed through the disarmament, demobilisation and reintegration (DDR) process, compared to 24,000 RUF. Of the 37,189 CDF, all but several thousand were Mende Kamajors; 34,890 were men, 1,996 were boys (younger than eighteen), 296 were women and seven were girls according to the National Commission for DDR. Christiana Solomon and Jeremy Ginifer, “Disarmament, Demobilisation and Reintegration in Sierra Leone”, Centre for International Cooperation and Security, University of Bradford, July 2008, p. 15.Hide Footnote Some joined to access weapons and other resources; others to settle scores. In joint operations with ECOMOG troops, they typically served as guides for troops unfamiliar with the territory or people. They also frequently were in the vanguard during attacks on rebel positions, with troops from the ECOMOG firing heavy artillery from behind. In advance of the rebel attack on Freetown in January 1999, ECOMOG airlifted Kamajors to help defend the capital.[fn]Crisis Group interview, former CDF commander, Freetown, 16 January 2017.Hide Footnote

The Kamajors’ reputed fearlessness [...] was matched by their brutality, especially when operating outside their home areas.

Neither national leaders nor ECOMOG (itself accused of complicity in Civil Defence Forces abuses) were willing or able to control such a large, decentralised, undisciplined and mostly untrained force.[fn]See for instance “Getting Away with Murder, Mutilation and Rape: New Testimony from Sierra Leone”, Human Rights Watch, July 1999.Hide Footnote The Kamajors’ reputed fearlessness – reinforced by initiation rites that were supposed to render fighters immune to bullets – was matched by their brutality, especially when operating outside their home areas.[fn]Sierra Leone Truth and Reconciliation Commission Report, Volume 3A, Chapter 4: Nature of the Conflict, pp. 541-542.Hide Footnote In larger cities such as Freetown and Bo, they robbed and harassed civilians, killing those suspected of collaborating with the enemy; in rural areas they were accused of committing massacres in supposedly pro-rebel villages.[fn]See Dalby, “In Search of the Kamajors”, op. cit. Also, “Sierra Leone: Most Serious Attacks in Months”, Human Rights Watch, 24 July 2001.Hide Footnote There lies in this a cautionary tale: the state’s willingness to empower civilians to fight on its behalf can trigger mass, unregulated recruitment, swelling a vigilante force beyond the state’s ability to oversee, let alone control, it.

2. A bitter legacy

At the end of the war, the government and international partners faced multiple imperatives: to disarm and demobilise the Kamajors alongside other combatants; recognise and reward their efforts; uphold justice and hold accountable those who committed abuses; and reconcile former enemies. Although the government took steps on all fronts, former Kamajors saw its limited support for reintegration as a sign of ingratitude and assumed the prosecution of their leaders was politically motivated.

The July 1999 Lomé peace accord soon was broken and fighting only died down after Britain dispatched 800 troops in May 2000 to stop a rebel advance on Freetown.[fn]In 1999, as domestic pressure grew on Nigerian President Olusegun Obasanjo to withdraw his troops from the ECOMOG mission – the major bulwark against the rebels’ advance – the international community pressed the Sierra Leonean government and the RUF to negotiate a peace agreement. The Lomé deal included a cessation of hostilities, disarmament and demobilisation programs, positions for the RUF in government, and amnesty and reintegration support for all fighters. Mohamed Gibril Sesay and Mohamed Suma, “Transitional Justice and DDR: The Case of Sierra Leone”, International Centre for Transitional Justice, June 2009, p. 10.Hide Footnote A year later, a disarmament, demobilisation and reintegration process got underway for more than 72,000 former combatants, including the Civil Defence Forces (CDF).[fn]DDR genuinely began in May 2001 after two previous unsuccessful attempts. Christiana Solomon and Jeremy Ginifer, “Disarmament, Demobilisation and Reintegration in Sierra Leone”, Centre for International Cooperation and Security, July 2008, pp. 8-11.Hide Footnote Most, incentivised by the promise of reintegration support, quit willingly. In January 2002, the government formally disbanded the CDF and banned all tribal militias. An estimated 20 per cent of CDF fighters were integrated into the security services. Others chose to continue their education. Most returned to their rural home areas or moved to provincial cities and tried to find work.[fn]Some high-ranking CDF members felt they were too senior to join the army as privates and so preferred to return to civilian life. Crisis Group interview, former CDF commander who became an officer in the Sierra Leone armed forces, Freetown, 11 January 2017.Hide Footnote

Many Kamajors [...] remain aggrieved, even bitter, that they did not receive the support to which they were entitled.

Government and donors paid less attention to reintegrating former fighters than to the disarmament and demobilisation phases; administrators acknowledge that vocational training courses were too short and did not fit economic needs. Many Kamajors, both leaders and foot soldiers, remain aggrieved, even bitter, that they did not receive the support to which they were entitled.[fn]Crisis Group interviews, former Kamajors, Freetown, Bo and Gerihun; former DDR administrator, Freetown, January 2017. The government and donors invested more in disarmament and demobilisation than reintegration partly because progress in the former was more measurable. Mohamed Gibril Sesay and Mohamed Suma, “Transitional Justice and DDR”, op. cit., p. 15.Hide Footnote

From 2002 to 2004, a Truth and Reconciliation Commission sought to heal the societal wounds caused by atrocities on all sides. Former Kamajors were among those who admitted their crimes, apologised to victims and asked the families of those killed for forgiveness. In 2002, the government and UN set up the Special Court for Sierra Leone to try those “bearing the greatest responsibility” for crimes against civilians and UN peacekeepers. Its prosecutor indicted thirteen people: nine Revolutionary United Front and Armed Forces Revolutionary Council rebel leaders plus then-Liberian president Charles Taylor (who backed their insurgency) and three CDF militia leaders, including Norman, who died in custody after undergoing medical treatment.[fn]In addition to Norman, the Special Court convicted two other CDF commanders, who served lengthy prison terms. In May 2008, an appeals chamber of the Special Court increased the original sentences of CDF leaders Allieu Kondewa and Moinina Fofana’s to twenty and fifteen years respectively. Special Court for Sierra Leone, Appeal Judgement, SCSL-04-14-A-829, 28 May 2008, p. 191.Hide Footnote Many Kamajors believe Norman’s indictment was designed to stop him from competing for the presidency and that his death at a military hospital in Senegal was no accident.[fn]“Special Report: Samuel Hinga Norman Dies”, Sierra Leone Trial Monitoring Program, UC Berkeley War Crimes Studies Center, 22 February 2007.Hide Footnote

Such suspicions reinforce the conviction among former Kamajors that the government failed to appreciate their sacrifices and ultimately betrayed them. Their leaders, especially Norman himself, had promised them recognition, including medals, and the transformation of the CDF into a reserve force, although these proposals never received cabinet approval. There remains little public recognition of the group’s contributions. A small monument next to the central roundabout in downtown Freetown bears a plaque reading: “To commemorate the work of the Civil Defence Force (CDF) in pursuit of peace and democracy in Sierra Leone, 1997-2002”.

Today, former Kamajors, especially in rural areas, still bear these grievances; the power shift to a northerner-dominated government since 2007 has compounded feelings of marginalisation in Mendeland. Still, the absence of a collective Kamajor voice and emergence of new political leaders and rivalries, including within the Mende-dominated Sierra Leone People’s Party, over time diluted the political significance of this perceived betrayal of the Kamajors.[fn]That said, the vice president’s successful attempts to bring Norman’s son onto his team suggest there may still be political currency in playing the Kamajor card in the 2018 elections.Hide Footnote

B. Arrow Boys of Teso in Uganda

The Lord’s Resistance Army (LRA), led by self-styled spirit-medium Joseph Kony, emerged in the late 1980s among disaffected ethnic Acholis in northern Uganda. It sparked an extraordinarily violent rebellion that would kill, mutilate and kidnap thousands of civilians in four countries over nearly three decades.[fn]On the LRA, see Crisis Group Africa Reports N°157, LRA: A Regional Strategy beyond Killing Kony, 28 April 2010; N°182, The Lord’s Resistance Army: End Game?, 17 November 2011.Hide Footnote The Ugandan army fought back, but could not or would not protect civilians from the LRA’s brutal attacks, prompting some to form vigilante or self-defence groups.[fn]The Amuka Boys were established in Lango sub-region, the Frontier Guards in Kitgum district, and the Arrow Boys of Teso sub-region (formerly a district) in the east. See map of Uganda in Appendix B.Hide Footnote Among the most effective were the Arrow Boys of Teso, a sub-region of eastern Uganda.[fn]See map in Appendix B.Hide Footnote With military backing and leadership (though minimal resources), local recruits – often led by ex-rebels who had once fought the central government – took up arms against the LRA in June 2003, driving it out of Teso by the end of that year.

Their success testified to their fighting ability and community support as well as their local leaders’ ability to secure national-level backing. Operating among their home communities under close political oversight by national and local leaders and a degree of military oversight by the national army, few Arrow Boys abused their power. However, because of a flawed demobilisation process, many Arrow Boys returned home without pay or lasting state support and grew resentful of the central government.

1. The “little army within the army”

The Teso Arrow Boys emerged in June 2003 in response to LRA attacks. Unlike other groups, they did not evolve from traditional tribal networks, such as the hunter societies that would become Sierra Leone’s vigilantes. They earned their name, according to a former commander, not because they shot arrows but “because they were like an arrow, which flies silently – like it knows where it is going”.[fn]Crisis Group interview, Musa Ecweru, former Arrow Boys Operational Commander, now a Member of Parliament and Minister of State for Relief, Disaster Preparedness and Refugees, Kampala, 23 January 2017.Hide Footnote Many of their leaders were former insurgents, who had honed their skills during the Uganda People’s Army (UPA) 1986-1992 uprising, as well as earlier rebellions.[fn]The Teso region had lent support to former President Milton Obote in the Bush War (1980-1986) against the National Resistance Army’s (NRA) rebellion led by Yoweri Museveni. After assuming power, Museveni remained suspicious of Teso’s political commitment to the new regime. The new army, seeking to entrench its dominance, tried to disarm militias in the area, originally formed to protect against cattle raiding from the neighbouring Karamoja region. This heavy-handed security approach and disbanding of militias – to which local communities turned for security when not offered by the government – motivated former militia members to form the Uganda People’s Army and recruit within the Teso region. Crisis Group interviews, former UPA members, Soroti and Kampala, January 2017. See also Ben Jones, “Remembering the Teso insurgency”, The Guardian, 24 February 2009.Hide Footnote “We had so many revolutions [in Uganda] that there were many ex-combatants in the villages”, said one former Arrow Boy field commander.[fn]Crisis Group interview, Aditu Abibu, Arrow Boys’ field commander, Soroti, 27 January 2017. Abibu also fought with the Uganda National Liberation Front against former President Idi Amin.Hide Footnote

After the Uganda People’s Army’s so-called Teso War ended, some ex-combatants were integrated into Anti-Stock Theft Units (ASTU) or Local Defence Units (LDU) to provide security against cattle raiders who repeatedly made sorties into Teso from the Karamoja region to the north east. Others simply returned to their villages.[fn]Ibid.Hide Footnote Thus the area had a pre-existing, albeit rudimentary, community defence structure. Local leaders initially reacted warily to the LRA’s arrival. Some preferred to let the group “pass through Teso unhindered”, fearing that confronting it would “endanger the lives of their people”.[fn]Crisis Group interview, Robert Adiama, former Arrow Boy counter-intelligence chief, Kampala, 24 January 2017.Hide Footnote Given the brutality of the government’s counter-insurgency operations just a decade earlier and the desire to avoid further conflict in Teso, initially there was limited enthusiasm for joining hostilities on either side.[fn]Crisis Group interviews, Paul Omer, former mayor of Soroti, via phone from Kampala, 30 January 2017; former Arrow Boys commanders, Soroti, January 2017.Hide Footnote

That changed when the LRA unleashed its violent campaign of child abduction in Teso; its methods convinced the community to mobilise. “We reacted as a tribe”, said the mayor of Soroti, Teso’s capital. “It was an issue of survival”.[fn]Crisis Group interview, Paul Omer, former mayor of Soroti, via phone from Kampala, 30 January 2017.Hide Footnote The emerging Arrow Boy leadership argued that the community itself must take the lead in opposing the rebels as army presence in the Eastern region was thin; troops were based in urban centres, unable to respond quickly to the LRA’s guerrilla tactics.[fn]Crisis Group interview, John Eresu, Kampala, 24 January 2017. Eresu, a former Arrow Boy commander and former army officer, was MP for a Teso constituency in 2003.Hide Footnote “A snake had entered our house”, said a local official and Arrow Boy officer. “You do not wait”.[fn]Crisis Group interview, Michael Bwalatum, Soroti, 29 January 2017. Bwalatum was the Arrow Boys’ administrative officer for welfare and Deputy Resident District Commissioner in Teso.Hide Footnote

Senior Teso political leaders – who notably included Musa Ecweru, a regional district commissioner, and Captain Mike Mukula, a former pilot who was then minister for health – held a meeting in early June 2003 to mobilise the community.[fn]Crisis Group interview, Musa Ecweru, Kampala, 23 January 2017. Ecweru, who left his post as a regional administrator to organise the Arrow Boys, served as the group’s operations coordinator. He previously represented the UPA in peace negotiations with the government. Since 2006, he has served as an MP for Teso and state minister for relief, disaster preparedness and refugees. Mike Mukula became the Arrow Boys’ chairman while serving as minister for health (2001-2006). He is now a successful businessman who runs, among other ventures, a private security company that employs former members of the force.Hide Footnote Radio stations called for recruits and local church networks relayed the message. The Anglican bishop of Soroti raised donations to pay volunteers. Using a few dozen arms supplied by the internal security agency, the Arrow Boys launched their first attack on 22 June, routing LRA rebels taken by surprise.[fn]Crisis Group interview, Musa Ecweru, Kampala, 23 January 2017.Hide Footnote

To survive future attacks and reprisals, Teso leaders needed to convince President Museveni to provide significant support. The decision involved risks for both sides. For the president, it implied giving weapons to former insurgents in a historically anti-government region.[fn]Crisis Group interview, Dr. Chris Dolan, Kampala, 23 January 2017. Dr. Dolan, an expert on the LRA, directs the Refugee Law Project at Makerere University.Hide Footnote For Teso politicians, it meant persuading local combatants to put aside their distrust of the army and accept its oversight.

In effect, [the Arrow Boys] formed “a little army within the army”.

Given the magnitude and immediacy of the LRA threat, however, neither side had much choice. The Uganda People’s Defence Force (the regular army) or UPDF was overstretched, lacking local intelligence, and reluctant to conduct anti-guerrilla operations in difficult terrain.[fn]On the importance of local collaboration in anti-LRA operations, see Crisis Group Africa Report N°157, LRA: A Regional Strategy Beyond Killing Kony, 28 April 2010, p. 6.Hide Footnote The Arrow Boys could not effectively protect their communities without the logistical support – especially weapons – only the army could supply. Museveni accepted the gamble, but to oversee the counter-insurgency campaign and make sure the Arrow Boys did not get out of hand he travelled regularly between Kampala and Teso.[fn]Crisis Group interview, Mike Mukula, Kampala, 20 January 2017.Hide Footnote For their part, local politicians set aside ethnic or regional resentments, assuring the government that the Arrow Boys would “assist” army troops rather than act independently. In effect, they formed “a little army within the army”.[fn]Crisis Group interviews, Musa Ecweru, Kampala, 23 January 2017; civil society representative, Kampala, 19 January 2017.Hide Footnote

The government distributed roughly 7,000 rifles to the Arrow Boys, who were organised as an auxiliary force divided into twelve battalions, each under the command of an army major. The estimated total size of the force was 9,000 including some unarmed members who focused on scouting or logistics roles, among them women.[fn]Crisis Group interviews, former Arrow Boy commanders, Soroti and Kampala, January 2017.Hide Footnote Relations with the army at times were fraught. The Teso combatants chafed under the army’s “formal way of doing things” and resisted demands they speak Kiswahili, the language used by soldiers, but the collaboration was militarily effective.[fn]Crisis Group interview, Michael Bwalatum, former Arrow Boys administrative officer, Soroti, 29 January 2017.Hide Footnote The Arrow Boys proved to be a highly motivated, mobile force that took the fight to the guerrillas, pursuing them on foot into the swamps of the Lake Kyoga basin. They harried the rebels relentlessly, a former army officer said, denying them the chance to rest and resupply.[fn]Crisis Group interview, John Eresu, Kampala, 24 January 2017.Hide Footnote Because they enjoyed the trust of local communities, the Arrow Boys provided the army with up-to-date intelligence, including through a network of village churches.[fn]Crisis Group interview, Bishop (retired) Bernard Obaikal, Soroti, 30 January 2017.Hide Footnote

Within the region, the force enjoyed overwhelming support for stopping rebel killings and kidnappings. There is little evidence that members abused civilians or engaged in criminal activity. “Crimes by the Arrow Boys against the community were very rare”, a former field commander said, though he admitted that “some of the boys were a bit lawless”.[fn]Crisis Group interview, Moxon Ekurit, Arrow Boy commander, Soroti, 28 January 2017.Hide Footnote Veterans of the force say discipline was closely monitored with infractions punished by their own commanders or by army courts-martial.[fn]Crisis Group interviews, Moxon Ekurit and Michael Bwalatum, Soroti, 28-29 January 2017.Hide Footnote

2. Flawed demobilisation

By the end of 2003 – only six months after local leaders met to plan community defence – the Arrow Boys had forced most LRA guerrillas out of Teso.[fn]The LRA attempted to return in 2005 when the army killed a well-known commander, initially reported to be Dominic Ongwen, now on trial before the International Criminal Court. See “LRA brigadier killed in Teso”, New Vision, 5 October 2005; “Ex-child soldier Dominic Ongwen denies war crimes at ICC trial”, The Guardian, 6 December 2016.Hide Footnote The force was then gradually demobilised and the last three battalions were disbanded in 2007. Some members simply “deserted”, returning to their villages as the LRA threat declined (though a former commander said they were quickly found and returned their weapons). The majority went through a formal process, which meant relinquishing their rifles and uniforms theoretically in return for payment. A small number joined the army, police or the Local Defence Units created mainly to repel Karamojong cattle raiders. Some former Arrow Boys eventually joined the large Ugandan army contingent in the African Union Mission in Somalia (AMISOM), a highly desirable posting given its salary and demobilisation payment.[fn]“UPDF recalls 2,000 reservists for Somalia duty”, Daily Monitor, 26 November 2012.Hide Footnote But despite their military success against the LRA, few met the educational requirements (a secondary education certificate) required to join the armed forces.[fn]Crisis Group interview, Michael Bwalatum, former Arrow Boys administrative officer, Soroti, 29 January 2017.Hide Footnote

Army officers beat some of those who requested compensation and sent them back to their villages empty handed.

Although each demobilised Arrow Boy was supposed to receive 840,000 Ugandan shillings (worth almost $500 in 2007 when the demobilisation process ended), former commanders say army officers stole much of the funding earmarked for this purpose. Nor did families of those killed in action receive promised “burial support”.[fn]Crisis Group interview, Robert Adiama, Kampala, 24 January 2017. Adiama wrote a report on the army’s alleged theft of demobilisation funds.Hide Footnote Instead, army officers beat some of those who requested compensation and sent them back to their villages empty handed.[fn]Crisis Group interview, Paul Omer, Mayor of Soroti, via phone from Kampala, 30 January 2017.Hide Footnote Local religious and political leaders have complained publicly about the government’s failure to offer the Arrow Boys adequate material or symbolic recognition for their service.[fn]Crisis Group interview, Bishop (retired) Bernard Obaikal, Soroti, 30 January 2017.Hide Footnote “I blame the government I serve for not rewarding [the Arrow Boys] with medals”, wrote the force’s ex-chairman, “and yet I see the government giving out medals to different groups across the country”.[fn]“Capt Mukula attacks government over Arrow Boys”, Daily Monitor, 5 January 2017.Hide Footnote

The absence of an effective demobilisation program and the government’s failure to properly acknowledge Arrow Boys’ services fuelled a strong sense of disillusionment with Museveni’s regime.[fn]Crisis Group interviews, former Arrow Boys, Kampala and Soroti, January 2017.Hide Footnote So far at least, however, this has not had a visible impact on political stability in the Teso region which has remained largely peaceful since the LRA left. (Karamojong cattle raiding also has declined due to a government disarmament operation in the region). There appears to be a broad sense that security in the region substantially improved – a point Museveni regularly stresses – and opposition political support does not appear to have coalesced around the Arrow Boys.

Although most of the rank-and-file Arrow Boys did not benefit significantly from their service, its leadership – particularly Musa Ecweru and Mike Mukula – were politically rewarded. Ecweru was promoted from regional district commissioner for Kasese in western Uganda to MP for Amuria (a Teso constituency), and also has served as state minister for disaster preparedness and refugees since 2006. Mukula was MP for Soroti municipality until 2016 and now serves as national vice chairman for the National Resistance Movement (NRM) Eastern Uganda. Ecweru reportedly also enjoys good relations with President Museveni, and campaigned on the same platform during the 2016 general election.[fn]Crisis Group interview, Aditu Abibu, Arrow Boys’ field commander, Soroti, 27 January 2017.Hide Footnote He reportedly provides money for former Arrow Boys – in particular to pay for funeral costs – even though such occasional patronage cannot compensate for the government’s failure to properly implement a demobilisation, disarmament and reintegration (DDR) program.[fn]Crisis Group interviews, Aditu Abibu and Michael Bwalatum, Soroti, 27 and 29 January 2017.Hide Footnote Although the 2016 election evidenced growing anti-Museveni sentiment in the region, Arrow Boys do not appear to be a major factor in in this.[fn]Electoral Commission of Uganda, Presidential Elections 2016: District Summary Report.Hide Footnote Electoral Commission of Uganda, Presidential Elections 2016: District Summary Report (

C. Zande Arrow Boys in South Sudan

By the mid-2000s, the Lord’s Resistance Army (LRA) largely had been pushed out of Uganda into neighbouring Democratic Republic of Congo (DRC), Central African Republic (CAR) and what in 2011 would become South Sudan. Pursued by the Ugandan army, which worked with its neighbours’ national forces, small groups of LRA fighters attacked unprotected villages to seize supplies, kidnap new recruits and then disappear back into the jungle. In South Sudan’s Equatorias region, some ethnic Zande communities (referred to collectively as the Azande) had in 2005 formed defence forces to repel ethnic Dinka pastoralists who drove cattle onto land they considered their own.[fn]The southern third of South Sudan, bordering Uganda, DRC and CAR, is known as the Equatorias. It comprises the former states of Western, Central and Eastern Equatoria. President Salva Kiir dissolved South Sudan’s ten states in 2015, replacing them with 28 new states. Western Equatoria was divided into Gbudwe, Maridi and Amadi. See map in Appendix C.Hide Footnote They took on the name previously used in Uganda: Arrow Boys. From 2008, the threat of LRA attack spurred the growth of Arrow Boy units.

These civilian forces proved most useful for reconnaissance and early warning. Mutual distrust between the Azande and the armed forces, rooted in longstanding ethnic and political tensions, hampered their effectiveness, however. It also ultimately drew the Zande Arrow Boys into the civil war that roiled South Sudan from 2013.[fn]See Crisis Group Africa Report N°236, South Sudan’s South: Conflict in the Equatorias, 25 May 2016.Hide Footnote As the LRA threat declined, the central government’s approach to the Arrow Boys – a mix of neglect and hostility toward a group that demanded to be armed, mobilised and paid, but not subject to central government control – helped fuel their transformation from self-defence groups into rebels.

1. Filling a security vacuum

Zande areas in the far south west of the country saw some of the lowest levels of fighting during Sudan’s Second Civil War (1983-2005) and thus became a natural refuge for millions of displaced persons. Many were Bor Dinka who fled with their cattle into the Equatorias following a 1991 massacre of more than 1,000. In the 1990s, the Sudan People’s Liberation Movement/Army (SPLM/A), the political and military group then leading the rebellion against the central government in Khartoum, captured Yambio and settled many displaced Dinka and wounded veterans in this relatively quiet backwater.

Until this period few Azande had joined the Sudan People’s Liberation Army (SPLA), the rebel force, in part because they saw it as a Dinka force. Their experience after being “liberated” by the SPLA further confirmed this belief: SPLM/A members, many of whom were Dinka, overruled local leaders, preferentially allocated land to Dinka civilians and Dinka’s cattle roamed over Zande farmers’ crops.[fn]Crisis Group interviews, Zande political leaders, civil society members and civilians, Juba, Addis Ababa, 2014-2016.Hide Footnote The Dinka saw themselves as civilians fleeing a brutal war that the Azande were lucky to have avoided,[fn]Crisis Group interviews, Dinka SPLM members whose families fled to the Equatorias, Juba, 2014.Hide Footnote but the Azande saw them as invaders, backed by SPLA guns. Differences in perception regarding who had fought for independence, suffered or sacrificed the most, and regarding who was entitled to what, continue to shape views of Azande, Dinka settlers and the Juba government.[fn]On the Equatorias during South Sudan’s struggle for independence, see Crisis Group Report, South Sudan’s South, op. cit., pp. 3-6.Hide Footnote

The Arrow Boys or Aparanga Aguanza[fn]Aparanga Aguanza means Arrow Boys. They are also sometimes known as the Home Guard. Other ethnic groups in the Equatorias also formed Arrow Boys, but the Azande were the most numerous. Crisis Group Report, South Sudan’s South, op. cit., p. 31.Hide Footnote first emerged as local defence forces in 2005 after the SPLM/A signed the Comprehensive Peace Accord (CPA) with the Sudanese government, creating a pathway to South Sudan’s full independence in 2011.[fn]The CPA created a semi-autonomous government based in Juba and provided for the referendum that allowed South Sudan to become fully independent in 2011. For more on divisions in South Sudan and the first Arrow Boys, see Crisis Group Report, South Sudan’s South, op. cit., pp. 9-10, 31.Hide Footnote With the war ending, the Azande – some of whom had fled to Congo or Uganda – mobilised to kick the Dinka and especially their cattle off land they regarded theirs.[fn]Adam Branch and Zachariah Cherian Mampilly, “Winning the War, but Losing the Peace? The Dilemma of SPLM/A Civil Administration and the Tasks Ahead”, The Journal of Modern African Studies, vol. 43, no. 1 (March 2005), pp. 1-20; Hélène Caux, “The Dinka head home with up to 1.5 million cattle in epic trek across South Sudan”, UNHCR, 8 December 2005.Hide Footnote

The LRA resorted to extreme violence, kidnapped civilians and forced them to fight members of their own community.

By the end of 2008, however, the Azande faced a more lethal enemy, the LRA. Following collapsed peace talks, the Ugandan army, with U.S. support, launched “Operation Lightening Thunder”, attacking LRA camps in DRC. Many LRA guerrillas escaped across the border into Western Equatoria.[fn]On Operation Lightening Thunder, see Crisis Group Africa Report N°157, LRA: A Regional Strategy Beyond Killing Kony, 28 April 2010, pp. 1-4, 12-13.Hide Footnote The LRA resorted to extreme violence, kidnapped civilians and forced them to fight members of their own community. Those who escaped often had to go through painful reconciliation processes to be welcomed back into their communities.

South Sudan’s army – still known as the SPLA – initially paid little attention to LRA guerrillas, whom they regarded as Uganda’s problem. Juba was preoccupied with asserting territorial control across the south and believed that another war with Khartoum was imminent. Although the legislature appropriated the equivalent of approximately $2 million to support the Arrow Boys, the latter say they never received it.[fn]“Conflict in Western Equatoria”, Human Security Baseline Assessment for Sudan and South Sudan, Small Arms Survey, 17 July 2016, p. 5.Hide Footnote This official neglect combined with the largely Dinka-led army’s apparent reluctance to protect their brethren deepened Zande distrust of the new SPLM-led government.[fn]Crisis Group Reports, The Lord’s Resistance Army, op. cit., p. 6; and South Sudan’s South, op. cit., p. 8.Hide Footnote

If the Zande Arrow Boys lacked significant national government support, they received help from church leaders, businessmen and Western Equatoria state officials. Two governors who were former SPLA officers – Colonel Patrick Zamoi (2005-2006) and Col. Joseph Bakosoro (2010-2015) – became important patrons of the Arrow Boys, mobilising them to defend their villages and to back Bakosoro’s gubernatorial campaign.[fn]Bakosoro, having failed to win the SPLM nomination, ran for governor in 2010 as an independent. See Crisis Group Report, South Sudan’s South, op. cit., p. 32; Crisis Group interviews, Zande politicians, Juba, 2016; Arrow Boy members, by telephone, 2015.Hide Footnote This support heightened tensions with national authorities. Relations became especially contentious between the leadership in Juba and Bakosoro, a popular Zande politician who sometimes referred to the Arrow Boys as “my” army.[fn]Crisis Group Report, South Sudan’s South, op. cit., p. 15; Crisis Group interviews, Zande intellectuals, Juba, January, February 2016; Equatorian expert, via Skype, February 2016.Hide Footnote

At the height of LRA attacks between 2008 and 2011, the government had reason to be satisfied with the Arrow Boys’ performance as a local defence force.

Still, at the height of LRA attacks between 2008 and 2011, the government had reason to be satisfied with the Arrow Boys’ performance as a local defence force. Most were only lightly armed – with hunting rifles, machetes, and sometimes bows and arrows – and thus presented no threat to central authority. But they had the advantage of mobility and surprise over their guerrilla opponents. Like local groups in Sierra Leone and Uganda, they knew the terrain, which allowed them to predict the LRA’s likely routes and to conduct night-time patrols, at times ambushing, capturing or killing LRA guerrillas. They were a trusted source of information for remote communities, both about the LRA’s whereabouts and its tactics, such as pretending to surrender to enter villages unchallenged.[fn]Crisis Group Report, LRA: A Regional Strategy, op. cit., p. 13. See also, John Norris, “Field Dispatch: The Arrow Boys of Southern Sudan”, Huffington Post, 12 May 2010. Norris was executive director of Enough, the anti-genocide project at the Center for American Progress.Hide Footnote They also provided valuable intelligence that helped security forces, particularly Uganda’s army, undertake targeted operations.

Mistrust still hampered relations with government forces, however. Reluctant to work with the SPLA, some Arrow Boys cooperated instead with the Ugandan army.[fn]Crisis Group Report, LRA: A Regional Strategy, op. cit., p. 8.Hide Footnote South Sudan’s army, in turn, was unwilling to coordinate with the Arrow Boys, though it allowed the vigilante forces to patrol in remote forested areas they could not reach due to lack of transport and communications equipment. International advisers to the U.S.-backed multinational counter-LRA forces helped bridge this divide between SPLA soldiers and Arrow Boys, providing equipment to the former and teaching them how to use information provided by the latter.[fn]Crisis Group interview, SPLA officer leading counter-LRA efforts during the period, Juba, 2015.Hide Footnote

Being an Arrow Boy was not full-time work, so most continued to farm and support their families. But they could mobilise quickly when necessary, communicating via mobile phones and, in the most remote areas, with drums.[fn]Crisis Group Report, LRA: A Regional Strategy, op. cit., p. 13, 17.Hide Footnote Much of the local population actively helped the Arrow Boys, either by donating supplies or by becoming active members: a 2013 survey in Ezo and Tambura counties found that four out of five respondents had provided them with food and half said either they or another household member had served with them.[fn]“JSRP Survey in Western Equatoria”, Justice and Security Research Programme, March 2014, pp. 34-38.Hide Footnote Because they were volunteers, deployed as needed, their numbers are hard to estimate, but according to a UN official, in 2008 each of Western Equatoria’s ten counties officially maintained approximately 2,000 Arrow Boys.[fn]UN official in Yambio, cited in “Conflict in Western Equatoria”, Small Arms Survey, op. cit., p. 6.Hide Footnote

After 2011, as the LRA threat receded, many Arrow Boys returned to full-time farming. Nonetheless, the Azande repeatedly demanded that the government formally recognise these local forces by arming, equipping and paying them. The Arrow Boys insisted on operating without national level control, however, and refused to join the army to avoid deployment outside their home region.[fn]Crisis Group interviews, Western Equatoria officials, 2014-2016.Hide Footnote Although Juba rejected their demands, some Arrow Boys remained active, implementing in several areas a parallel justice system for small disputes.[fn]“JSRP survey”, op. cit., pp. 34-38.Hide Footnote The above-mentioned 2013 survey found that nearly 85 per cent of respondents trusted Arrow Boys for dispute resolution, more than those who trusted local chiefs, elders, the church or the SPLA. In 2010, the Arrow Boys once more got involved in an ethnic conflict, joining a state government-led campaign to forcibly expel nomadic Mbororo cattle herders, a violent effort that reportedly involved violations of both international and national human rights law.[fn]“Western Equatoria governor tells Ambororo pastoralists to leave Zande land”, Sudan Tribune, 8 October 2010.Hide Footnote

2. Entangled in civil war

In 2013, two years after South Sudan’s independence, civil war broke out again, this time between forces aligned with President Salva Kiir (a Dinka) and those associated with then-Vice President Riek Machar (a Nuer). The Azande initially did not get involved in the dispute, but the conflict revived old resentments. Violence in predominantly Dinka areas once more displaced cattle herders into Western Equatoria. As tensions grew and tit-for-tat violence escalated, the Azande perceived the government as supporting the Dinka. Juba, meanwhile, interpreted the Azande’s lacklustre response to its appeal for SPLA recruits as disloyalty. It also distrusted the region’s popular governor, Bakosoro, the Arrow Boy patron, who continued to use the Arrow Boys to further his own political goals. The situation came to a head in September 2015 when Kiir removed Bakosoro from office and the Arrow Boys entered into open rebellion.

Despite being motivated by essentially local grievances, the Arrow Boys inexorably were drawn into the civil war: like the Kamajors in Sierra Leone, they went on the offensive, only in this case against government troops. Most either joined the insurgent Sudan People’s Liberation Movement/Army-In Opposition or formed their own rebel groups, while the remainder persisted essentially as local forces dedicated to protecting their communities. A series of battlefield losses led many Arrow Boys to disperse; their large-scale rebellion effectively collapsed. The Arrow Boys’ local support also eroded, both because the SPLA retaliated by abusing Zande civilians and because groups calling themselves Arrow Boys began to operate as criminal gangs, robbing, attacking and raping civilians.[fn]Crisis Group interviews, Western Equatoria church leaders, Juba, 2015-2017.Hide Footnote

In mid-2017, some Arrow Boys still were at war and ambushed government vehicles or blocked roads in forested areas. Yet most of Zandeland was in a “negative peace”: there was little fighting but the conflict remained unresolved. One large Arrow Boy group signed a peace agreement with Juba but it remained unimplemented. Many Arrow Boys are returning to their communities where they encounter a lukewarm welcome. Such is the extent of their loss of status that churches in some communities are organising reconciliation processes for returned Arrow Boys akin to those used for LRA escapees.[fn]Crisis Group interviews, Western Equatoria church leaders, Juba, 2016-2017.Hide Footnote

D. Nigeria’s Civilian Joint Task Force

The radical Islamist movement known as Boko Haram launched its insurgency in 2009 from the city of Maiduguri, the capital of Borno state in north-eastern Nigeria. From there it spread to the border areas of Chad, Niger and Cameroon. Unemployed urban youths made up most of the original movement, led by a charismatic young preacher named Mohammed Yusuf, who rejected secular authority and sought to establish a caliphate.[fn]Crisis Group Africa Report N°168, Northern Nigeria: Background to Conflict, p. 18.Hide Footnote A brutal 2009 crackdown by Nigeria’s security forces in Maiduguri – including Yusuf’s death while in police custody – drove the movement underground, fuelling an insurgency that in time would spread throughout the Lake Chad basin.[fn]Crisis Group Africa Report N°216, Curbing Violence in Nigeria (II): The Boko Haram Insurgency, 3 April 2014, pp. 7-14.Hide Footnote The group’s tactics have varied over time and place; it has terrorised the region with both suicide bombings in larger cities – sometimes well beyond the north east and up to the federal capital, Abuja – and guerrilla attacks on rural towns and villages, and has conducted mass abductions of youths and women, including schoolgirls.[fn]Crisis Group Africa Report N°242, Nigeria: Women and the Boko Haram Insurgency, 5 December 2016, p. 7. “‘Our Job is to Shoot, Slaughter and Kill’: Boko Haram’s Reign of Terror in North East Nigeria”, Amnesty International, 14 April 2015.Hide Footnote In response, citizens organised vigilante groups to protect themselves both from Boko Haram and the government’s often brutal counter-insurgency campaigns. While these groups have helped the police and military launch more targeted, effective operations, they also at times abused their authority.

1. From vigilantes to civilian task force

After 2009, Boko Haram attacked security forces as well as a wide range of civilian targets, including clerics, local politicians, neighbourhood chiefs and students attending secular, state-run schools.[fn]On the escalation and spread of Boko Haram attacks in north-eastern Nigeria see Crisis Group Report, Curbing Violence, op. cit., pp. 14-18.Hide Footnote In early 2013, according to local accounts, several residents decided that citizens of Maiduguri should organise to defend themselves.[fn]See, for example, “Nigeria: Civilian JTF – Unsung Heroes of the Boko Haram War”, This Day, 4 October 2015.Hide Footnote They started by seeking out, attacking and killing Boko Haram members. By June of that year, roughly 500 vigilantes were manning checkpoints, armed only with sticks and machetes, to spot and eliminate Boko Haram members moving about in, or trying to escape from, Maiduguri. They called themselves the Civilian Joint Task Force (CJTF), a name chosen to suggest they were a counterpart to the government’s Joint Task Force (JTF) of army, air force, police and other security units assigned to fight Boko Haram in Borno state.[fn]The JTF was replaced in August 2013 by the army’s newly created 7th infantry division.Hide Footnote

The vigilantes were protecting themselves from a dual threat: both from Boko Haram and from government security forces.

The vigilantes were protecting themselves from a dual threat: both from Boko Haram and from government security forces, which were inflicting collective punishment on communities suspected of harbouring militants, sometimes setting fire to houses and shops or randomly arresting – and in some instances, executing – passers-by.[fn]“Spiralling Violence: Boko Haram Attacks and Security Force Abuses in Nigeria”, Human Rights Watch, 2012, p. 59.Hide Footnote Citizens of Maiduguri also may have hoped to ease the state of emergency imposed in May 2013, which included suspension of phone services, a measure that largely crippled commerce and communication across the region.[fn]Crisis Group Africa Report N°244, Watchmen of Lake Chad: Vigilante Groups Fighting Boko Haram, 23 February 2017, pp. 4-5.Hide Footnote

Soon after its emergence, security services and civilian authorities became closely involved in the Civilian Joint Task Force’s organisation, management and operations. The army-led Joint Task Force quickly recognised the vigilantes’ potential. With the help of local and traditional authorities, it organised them according to its own command structure, establishing a CJTF unit for each of Maiduguri’s ten security sectors. Joint Task Force officers helped select vigilante leaders and Borno state officials became involved in management roles. Beginning in September 2013, the state government formally incorporated the CJTF under the Borno Youths Empowerment Scheme (BOYES) and selected around 1,850 young men – a small portion of total CJTF membership – for basic combat training.[fn]The government announced plans to train up to 6,000 young men but then scaled back, apparently concerned about training so many potentially uncontrollable people. Crisis Group interviews, CJTF leaders and BOYES members, Maiduguri, 18 October 2016, January 2017.Hide Footnote The state gave them uniforms, cars, identification documents and a stipend; the army subsequently provided standard military training to some 200 additional members to create a “CJTF Special Force” for front-line operations. Those selected for military training went through a vetting process, including background checks and medical screening.[fn]For more on BOYES, see Crisis Group Africa Report, Watchmen, op. cit., pp. 9-10.Hide Footnote Usually only sector commanders carried modern weapons, although the army provided members with assault rifles for specific operations.[fn]Crisis Group interviews, CJTF leader, Adamawa, 31 October 2016; CJTF, Maiduguri, January 2017.Hide Footnote

In mid-2013, CJTF members from Maiduguri began accompanying the army outside the city, working with them to form units in locations that had been under attack or recovered from Boko Haram. Most rural units had only traditional weapons, such as spears, bows and arrows or locally manufactured shotguns. The force also spread to other north-eastern states. In Adamawa state, the Kanuri minority, mostly traders in the state capital, Yola, formed its own 300-man CJTF in March 2013.[fn]Because many Boko Haram leaders are Kanuri, a large ethnic group living throughout the Lake Chad basin, this community frequently is suspected of supporting the insurgents.Hide Footnote When in late 2014 Boko Haram threatened Yola, hunter brotherhoods from various communities and ethnicities mobilised in response and were strongly supported by state authorities and local elites.[fn]Crisis Group interviews, state government officials and hunters’ association leader, Yola, November 2016.Hide Footnote

Like other civilian defence groups, these units carried out intelligence and surveillance missions, patrolled roads and manned checkpoints. Their local knowledge allowed them to identify and vet newcomers spotted in public spaces vulnerable to attack, such as mosques and markets.[fn]Crisis Group Report, Watchmen, op. cit., pp. 9-10.Hide Footnote They monitored and provided security for communities displaced by the conflict, including the almost two million people in IDP camps in north-eastern Nigeria.[fn]1.9 million people remained internally displaced in northern Nigeria in June 2017 and over 200,000 had been forced to flee to Cameroon, Chad and Niger. UN Office for the Coordination of Humanitarian Affairs (www.unocha.org/nigeria/about-ocha-nigeria/about-crisis).Hide Footnote Women have participated in patrols and, occasionally, combat. They also are used to search other women, a job that is especially important given cultural sensitivities about men searching women and Boko Haram’s use of women both as fighters and suicide bombers.[fn]On women in the insurgency and the CJTF, see Crisis Group Africa Report N°242, Nigeria: Women and the Boko Haram Insurgency, 5 December 2016.Hide Footnote

But members of the CJTF have gone further. They have acted as police auxiliaries, arresting suspects and participating in interrogations. The military at times deployed them in long-distance operations, mixing vigilantes familiar with local conditions and outsiders. CJTF members can remain in liberated towns to support local civilian forces. Authorities occasionally used the CJTF for autonomous operations, such as patrolling corridors used by Boko Haram to attack villages in Chibok.[fn]Crisis Group interviews, CJTF, Maiduguri, January 2017.Hide Footnote

2. Mixed record

Benefits of these citizens’ task forces are manifold. Their knowledge of local languages and terrain, both physical and social, helps security forces better target their operations. Because local civilians have a “better sense of the normal and the abnormal”, they can detect threats, such as potential suicide bombers.[fn]Crisis Group interview, security expert, Maiduguri, 15 January 2017.Hide Footnote They can serve as trusted links between security forces and locals. Affiliation with the CJTF can also protect its members from the army and police.

But dangers exist for both task force members and their communities. In Borno state, where the large majority of CJTF casualties have been recorded, 680 CJTF members were killed between 2014 and mid-2017.[fn]‘‘Boko Haram kills 680 Civilian JTF members in three years’’, Premium Times, 9 July 2017.Hide Footnote Cities and towns that formed citizen security groups also paid a price, as Boko Haram targeted traditional chiefs and other CJTF supporters. In June 2013, the group declared “all-out war” on the youth of Maiduguri and Damaturu “because [they] have formed an alliance with the Nigerian military and police to fight our brethren”.[fn]Audio clip, 18 June 2013, from Boko Haram spokesman Abu Zinnira quoted in “Civilian vigilante groups increase dangers in northeastern Nigeria”, IRIN, 12 December 2013. Boko Haram for a while set up roadblocks where it executed any male traveller from Maiduguri, to punish the city. Crisis Group interview, Maiduguri, January 2017.Hide Footnote Casualties peaked in 2013-2014, due largely to such retribution.[fn]For examples of such reprisals, see “Boko Haram weekend killing spree leaves at least 40 dead in Borno villages”, Information Nigeria, 29 July 2013; “Au Nigeria, ‘Boko Haram élimine des villages entiers suspectés d’avoir collaborate avec le pouvoir’”, Le Monde, 18 March 2014; “Boko Haram suicide bomber attack home of Civilian JTF commander”, Sahara Reporters, 25 January 2017.Hide Footnote

There are reports of vigilantes exploiting their privileged status and relative impunity for criminal purposes.

Some task force members also have exploited their positions for revenge and profit. Few CJTF members receive a stipend; most depend on haphazard support from local authorities, politicians or business people. Others reportedly share with security forces the spoils captured from Boko Haram or receive a portion of the aid provided to IDP camps.[fn]Crisis Group Report, Watchmen, op. cit., p. 12.Hide Footnote There are reports of vigilantes exploiting their privileged status and relative impunity for criminal purposes, including small-scale drug trafficking and resale of stolen goods. Other activities are akin to protection rackets, such as when vigilantes request “donations” at checkpoints or impose a form of taxation on local communities.

Even more troubling are reports of CJTF atrocities. Particularly during their early years and in the heat of the fight to expel Boko Haram from Maiduguri, vigilantes engaged in summary executions, often in collusion with the military. The CJTF reportedly burned alive several Boko Haram suspects in 2013. In one of the most notorious cases, task force members and soldiers rounded up hundreds of prisoners who had escaped from a military detention centre in Maiduguri before killing them. Vigilantes in a town in southern Borno reportedly paraded with the heads of 40 alleged Boko Haram militants on pikes.[fn]“Boko Haram hunters burn suspect alive in Maiduguri”, Daily Trust, 25 July 2013; “Nigeria. Les crimes de masse de Boko Haram”, International Federation for Human Rights, 10 February 2015, p. 9; “41 Boko Haram members beheaded in Biu after failed attack”, Daily Post, 31 October 2014; “Stars on their Shoulders, Blood on their Hands. War Crimes Committed by the Nigerian Military”, Amnesty International, June 2015. For another case, “How I escaped death in Maiduguri – Ex-Borno commissioner”, Vanguard, 25 March 2014. For a video account, see “Nigeria’s hidden war: Channel 4 dispatches”, Channel 4, 18 August 2014. CJTF officials denied their members were involved. Crisis Group interviews, January 2017; Maiduguri, 18 October 2016.Hide Footnote

Over the long run, as the threat from Boko Haram declines, the political risk posed by the CJTF could well increase. Some of its leaders make clear they expect to be rewarded with jobs or other compensation. As a federal government response, about 250 CJTF members were absorbed into the army in 2016.[fn]‘‘250 former Civilian-JTF members join Nigerian Army’’, Premium Times, 22 July 2016.Hide Footnote Another 120 were recruited by the domestic intelligence agency, Department of State Services, while 40 were enlisted by the air force.[fn]Figures provided by CJTF’s legal adviser, Jibril Gunda, at Summit on Security and Governance in the North-East, organised by CLEEN Foundation, Nigeria Stabilisation and Reconciliation Program (NSRP) and Ford Foundation, Gombe, Gombe state, 29 June 2015.Hide Footnote More recently, in May 2017, Labour Minister Chris Ngige said the federal government plans to train CJTF members in various vocational skills at the North-East Zone Skills Upgrading Training Centre in Bauchi, as reward for fighting Boko Haram.[fn]‘‘FG to compensate Civilian JTF for role in fighting insurgency’’, PM News, 23 May 2017.Hide Footnote

As the threat from Boko Haram declines, the political risk posed by the CJTF could well increase.

The Borno Youths Empowerment Scheme program offers professional training to the CJTF, but it benefits only a fraction of the whole group, estimated to number between 15,000 and 20,000 in Borno state alone.[fn]Estimating the size of the CJTF’s membership is difficult due to irregular recruitment and demobilisation. In mid-2014, CJTF leaders claimed to have recruited 45,000 members. In late 2016, they said between 22,000 and 26,000 members remained and, in January 2017, 15,000. Crisis Group interviews, CJTF leaders, Maiduguri, October 2016 and January 2017; “Nigeria – the community turns against Boko Haram”, IRIN, 11 August 2014.Hide Footnote The Borno State Vigilante and Youth Empowerment Agency Law approved in May 2015 is supposed to facilitate job creation for youth, particularly targeting CJTF members. And the governor of Borno has promised to provide about 20,000 jobs to former task force members.[fn]Crisis Group interview, CJTF leaders, Maiduguri, 18 October 2016; “They’re defeating Boko Haram but are they Nigeria’s next security threat?”, IRIN, 22 August 2016.Hide Footnote So far, these programs and promises have not had much success. Should they not materialise, Nigeria may be left with another angry armed group in the troubled north east. Some CJTF members allegedly now work for state politicians, who are known to employ thugs to attack opponents. Others may move further into extortion, drug trafficking and other organised criminal activity.

III. Lessons from the Past

Vigilante groups in violent conflict pose a dilemma: they can protect civilians and help regular forces overcome deadly insurgent groups but also risk attacking rival communities or preying on towns and villages they are supposed to protect. In the worst cases – the Kamajors of Sierra Leone or the Arrow Boys of South Sudan – local vigilante groups can end up as powerful ethnic militias or outright insurgents that help drag the country further into civil war.

This risk is inherent to the circumstances in which vigilantes are most likely to emerge – where weak states cannot on their own confront armed groups. Insofar as vigilante groups tend to form within local communities, members typically share the same ethnic or political identity, collective interests and threat perceptions. As a result, they are prone to have agendas that diverge from that of the central state or even bring them into confrontation with it. States too weak to protect communities from insurgents more often than not will be too weak to prevent vigilantes from using their power to pursue those agendas or abusing civilians.

This outcome is not preordained. As the case studies suggest, certain factors and behaviour by central governments can make such an evolution more or less likely.

A. Guarding against Mission Creep

Whether vigilante groups adhere to their original community protection and counter-insurgency roles or morph into ethnic militias and insurgent groups hinges in large part on local leaders’ agendas and relative autonomy from – or alignment with – national governments. To ward against vigilantes veering dangerously away from their original purpose, national governments would do well to engage local leaders with influence over vigilantes as they emerge, including traditional and religious authorities and business elite, with the aim of settling on finite, mutually acceptable objectives within an overarching national counter-insurgency strategy. Central states need to persuade vigilante leaders that adhering to these goals will benefit them and their communities, both through immediate security gains and further down the line in the post-conflict political settlement.

The contrasting approaches of governments in Uganda and South Sudan and their outcomes illustrate the need for this close political engagement. The Teso Arrow Boys of Uganda stand out by and large for having remained focused on their initial objective: protecting communities from the LRA and expelling insurgents from their area. They only diverged from this mission in fending off cattle raids by neighbouring Karamojong. This outcome stemmed largely from the willingness of both Museveni’s government and Teso leaders to agree on the Arrow Boys’ role within the broader counter-LRA campaign and how they would work with the national army. Museveni’s government and army were keen to restrict Arrow Boys’ mandate and operations and Teso politicians and other local leaders saw no advantage in turning the Arrow Boys into a new rebel group, despite lingering anti-Museveni sentiment in the area. Instead they sought political gain for themselves and their constituencies by nurturing relationships with Museveni and securing influential positions within his regime.

The [Teso Arrow Boys of Uganda] achieved their narrowly circumscribed objectives in short order and demobilised in the following few years.

With the army and other vigilante groups confronting the LRA elsewhere, the Teso Arrow Boys were not deployed beyond their region. Within Teso, governance systems were well established, so the Arrow Boys did not have the opportunity to expand their mandate into policing or dispute resolution roles. Thus, under strong local and national political oversight, with sufficient but cautious military support and significant local legitimacy, the Arrow Boys achieved their narrowly circumscribed objectives in short order and demobilised in the following few years.

In South Sudan, in contrast, ethno-political rivalry between the Dinka and the Azande meant the government treated the Arrow Boys with neglect and, eventually, hostility. Feeling abandoned by the central state and responding to community demands, the local forces diversified and expanded their security and governance roles, entrenching their positions of authority in local communities, and later siding with rebel factions against perceived Dinka aggression.

The government, which considered the LRA threat essentially Uganda’s problem, never fully backed the Arrow Boys, politically or militarily. This frustrated the Zande community and politicians, who unsuccessfully lobbied the government to arm and equip them, much like regular soldiers, even as they insisted on remaining independent. Largely free from national political oversight and endorsed by local traditional and religious leaders, the Arrow Boys expanded their roles and became increasingly autonomous of central state authority. They mobilised against Mbororo cattle herders and resolved local disputes, thus further establishing themselves as providers of security and governance. When a new influx of Dinka herders threatened Zande livelihoods with the backing of government forces, the Azande saw the soldiers as invaders and aligned themselves with pre-existing rebel groups.

Central governments keen to avoid the South Sudan scenario should strive to set vigilantes’ operations within a broader political bargain with local leaders that offers incentives for both sides to restrict vigilantes’ mandate. Through early and persistent engagement with vigilante representatives and influential community leaders such as religious figures and businesspeople, central states should aim to persuade vigilante leaders that they can best serve their individual and community interests by aligning vigilantes’ objectives with the state’s overall counter-insurgency strategy. Close oversight by national and local political leaders throughout the vigilantes’ mobilisation is critical to ensure they remain committed to mutually acceptable objectives. This also will make it possible to reassess the scope and intensity of vigilantes’ activities as the insurgency evolves.

In this context, international donors and partners would be best served by working with state authorities, helping them manage relations with vigilante groups, cautioning against the pitfalls of neglect, counterproductive repression and unfettered support. Likewise, they should avoid providing direct support to vigilantes, lest they weaken national authorities’ bargaining position.

B. Curbing Abuses

As the cases suggest, vigilante groups more often than not are guilty of committing egregious abuses, preying on civilians and becoming involved in illicit activities.[fn]This trend is not exclusive to sub-Saharan Africa. In Libya, Crisis Group found that the internationally-recognised Government of National Accord (GNA) has relied on local non-state armed groups in Tripoli for its own security, including the Rada Force, Tripoli Revolutionaries Brigade and Ghnewa Brigade. It also has sought to co-opt local armed groups into the coast guard and border guards. The GNA is struggling to control such groups, which benefit from its resources and legal cover. GNA recognition has emboldened them and encouraged their involvement in abuses, including kidnappings, arbitrary detentions and summary executions as well as illicit economic activity. In August 2017, the Tripoli Revolutionaries Brigade arrested and detained a former prime minister. See “Final report of the Panel of Experts on Libya established pursuant to resolution 1973 (2011)”, S/2017/466, 1 June 2017, pp. 2, 9, 41-42. “Libya’s coast guard abuses migrants despite E.U. funding and training”, Washington Post, 11 July 2017. In Algeria, to counter Islamist insurgents in the country’s civil war (1991-2002) the government from 1995 on backed local civilian forces, known as gardes communales or patriotes, but did not prevent them from committing deliberate and arbitrary killings. See “Algeria – Fear and Silence: A Hidden Human Rights Crisis”, Amnesty International, November 1996, pp. 17-19. In Colombia, between 1997 and 2006 the United Self-Defence Forces of Colombia, a national umbrella organisation for multiple local defence groups, while claiming to protect civilians from guerrillas, took territorial control and in places stepped into governance roles. Groups were responsible for major human rights violations and fuelled the cocaine trade. Crisis Group Latin America Reports N°11, War and Drugs in Colombia, 27 January 2005, pp. 13-16; and N°20, Colombia’s New Armed Groups, 10 May 2007, pp. 3-5.Hide Footnote Contrasting dynamics in Uganda and Sierra Leone offer insights into what factors and policies enable or reduce such tendencies. The Arrow Boys in Teso region committed limited abuses, deterred by the threat of internal discipline, military court-martial and shaming by their home communities. In contrast, the central government in Sierra Leone lent the vigilante group unguarded support, exerting insufficient oversight to stop tens of thousands of fighters from wreaking havoc among civilian communities.

Kamajors treated suspected rebel sympathisers and other civilians who resisted them brutally. This sense of being above the law stemmed largely from the Kamajors’ self-identification as defenders not just of Mendes in the south and east, but also of the central state which was then ruled by the Mende-dominated Sierra Leone People’s Party (SLPP). The government fostered this sense of national responsibility as it used the Kamajors to defend against the rebels and, with Mende figures sympathetic to the Kamajors at the heart of state structures (the deputy defence minister and vice president were early instigators), it maintained the flow of resources. In short, the Kamajors came to view themselves – and to be treated as – a substitute for the mostly defunct national army.

With the state administration in tatters, the embattled government exerted scant political or military oversight and enjoyed little control. The Kamajors’ military-type hierarchy was more honorific than functional and ranks were often self-assigned. Operational authority, therefore, fell mostly in the hands of battlefield commanders who, in the name of defending the elected government, accumulated weapons, often directed their forces for personal gain and failed to prevent the rank and file from committing atrocities. Operating under the Civil Defence Forces’ national mandate, Kamajor units left their home areas to fight in other regions where they could commit abuses with less fear of being identified.

While the Kamajors remained loyal to the regime out of ethnic solidarity, local commanders and foot soldiers used extreme violence in pursuit of self-serving agendas.

National leaders also struggled to control the Kamajors because of their large numbers. The Teso Arrow Boys peaked at about 9,000 members, mobilised for only four years. In contrast, the Kamajors, which began to organise in the early 1990s, mushroomed with state encouragement to over 30,000 members by war’s end in 2002. Economic incentives and social pressures fuelled largely unregulated recruitment. Members gained access to state resources and weapons, while some made a business of administering initiation rites. The Kamajors’ social status as community defenders also made it unacceptable in some places for men not to join. An international NGO worker who spent time with the Kamajors during the war said membership was “a job opportunity” in an organisation akin to “a social movement”.[fn]Crisis Group telephone interview, international NGO worker, 5 January 2017.Hide Footnote Government supplies, though significant, were still not enough for such a large number of combatants, who looted and extorted money from civilians. Thus, while the Kamajors remained loyal to the regime out of ethnic solidarity, local commanders and foot soldiers used extreme violence in pursuit of self-serving agendas.

To avoid this turn of events, central governments, with international support, can take several mitigating measures. Building on a political foundation of shared counter-insurgency goals, national leaders should encourage vigilante leaders and local community representatives to vet recruits more carefully. The larger the number of such recruits, the harder they are to control; recruitment thus could be capped at levels commensurate with state and local leaders’ capacity to oversee their activities.

Central governments could go so far as to insist that vigilantes operate only in their home areas – or, if displaced, among their own communities –, thereby reducing their contact with other ethnic groups and deterring abuses. Offenders will be more easily identified and shamed among their own people, facing potentially long-lasting consequences. As a further preventive measure, state authorities should, where possible, supply and equip vigilantes, reducing the risk that they might feel justified taking provisions and equipment by force from civilians or international aid organisations.

To hold offenders to account, central governments should advise vigilante and local leaders to establish their own codes of conduct and publicise them widely, including via radio. They also should establish their own disciplinary bodies to enforce rules of behaviour. In general, internal disciplinary processes are preferable to punishment by the national army, which risks opening rifts between vigilantes and regular soldiers. Central states and international partners also should encourage civil society and non-governmental organisations to conduct independent reporting on abuses and publicise their findings.

C. Balancing Security and Preservation of Central State Authority

Even if governments and donors take steps to ward against mission creep and abuses, empowering vigilantes has the potential of undermining central authority and tipping the power balance toward non-state armed actors. This is all the more likely when outside parties work in tandem with such actors, thereby affording them international legitimacy. In such cases, particularly if those outside parties act without the state’s consent, they risk prioritising short-term expediency over long-term state-building goals.[fn]In the Middle East in particular, Western powers have backed non-state armed groups fighting jihadist groups, at times undermining state unity and planting the seeds of new rifts. This was the case in Iraq, as illustrated below. See also Crisis Group Middle East Report N°74, Iraq after the Surge I: The New Sunni Landscape, 30 April 2008, p. ii. Likewise, U.S. support for the Kurdish-dominated Syrian Democratic Forces against the Islamic State (ISIS) in northern Syria presents its own sets of risks due to tensions with local Arab communities and with Turkey; in this instance, the Kurdish faction’s agenda – which goes beyond defeating ISIS – puts them at odds with domestic and foreign constituencies. It also could put them at odds with the regime. See Crisis Group Middle East Briefing N°53, Fighting ISIS: The Road to and beyond Raqqa, 28 April 2017; “The U.S. joins the Turkey-PKK fight in northern Syria”, Crisis Group op-ed, Middle East Eye, 12 May 2017.Hide Footnote For that reason, traditional counter-insurgency models often cast supporting vigilantes as a policy of last resort because it runs counter to the state’s monopoly over the legitimate use of violence.[fn]Critics of U.S. support for Sunni tribal armed groups in Iraq argued the strategy fomented “tribalism, warlordism, and sectarianism” and put short-term security gains above the country’s long-term stability and unity. See Steven Simon, “The Price of the Surge”, Foreign Affairs, May/June 2008. Counter-insurgency theorists have warned against entering short-term alliances with local forces “especially if [the latter] are likely to behave badly, change sides, or continue to call in coalition support” after the end of direct cooperation. See John Mackinlay and Alison Al-Baddawy, “Rethinking Counterinsurgency”, RAND Corporation, 2008, p. 59.Hide Footnote

Yet the proliferation of non-state armed actors in the context of deficient state security forces has forced a re-evaluation. As academic experts have noted, the notion of a state monopoly over the use of force often is divorced from reality; the truth is closer to an oligopoly.[fn]See Ariel Ahram, “Learning to live with militias: Toward a critical policy on state frailty”, Journal of Intervention and Statebuilding, vol. 5, no. 2, June 2011.Hide Footnote The challenge is how to manage such an arrangement when the state faces hostile insurgent groups and cannot provide security without relying on allied militias or where outside parties feel threatened by a terrorist-qua-insurgent group and therefore subcontract security duties to an allied militia group. In such instances, the urgent need to address the security menace can take precedence over the longer-term goal of state-building. A political order undergirded by a network of non-state actors and local strongmen hardly is optimal for building effective national institutions. But in fragile states facing civil conflict, such an imperfect order can be the lesser of two evils.

Governments should view partnering with vigilantes [...] as an opportunity to pursue the long-term objective of bolstering state legitimacy at the local level.

There are ways for the state to limit long-term damage, both while cooperating with vigilantes and after the insurgent threat subsides. Where possible, governments should view partnering with vigilantes not as a stop-gap or temporary alliance of convenience, but as an opportunity to pursue the long-term objective of bolstering state legitimacy at the local level. As the case studies illustrate, the state should cooperate closely with vigilante leaders and local elites in this regard, ensuring that all work together to oversee the actions of the vigilante groups, effectively manage vigilantes’ expectations and recognise their efforts.

In Uganda, Museveni turned the Teso Arrow Boys’ success to his advantage by elevating local Teso politicians into influential positions in his regime. After the LRA’s rout, they continued to represent their community’s interests so that a once hostile, potentially rebellious area had a stake in maintaining the status quo. Applying such a long-term strategic lens, governments and their partners should plan well in advance how they will manage vigilante groups after the insurgency recedes.

D. Planning for the Day After

Without a workable plan for managing vigilantes after the insurgency ends, governments face yet another risk: that vigilantes and their communities feel they have been used and abandoned. That threatens to alienate unemployed youth vulnerable to recruitment into anti-state factions, criminal gangs or radical groups. For example, because South Sudan failed to disband or formalise the Zande Arrow Boys after the LRA threat declined, they were able to join rebel ranks years later when they felt their community was threatened. As the Boko Haram threat wanes, Nigeria likewise is faced with the challenge of preventing members of the increasingly redundant Civilian Joint Task Force from turning to crime.[fn]See “Nigeria wakes up to its growing vigilante problem”, IRIN, 9 May 2017.Hide Footnote

As with other armed groups, disbanding vigilantes is likely to require a comprehensive disarmament, demobilisation and reintegration (DDR) process. In Sierra Leone and Teso, DDR processes were intended to offer vigilantes incentives to return to civilian life and turn over their weapons, reducing the number of arms in circulation. But both processes fell short, leaving vigilantes and their communities bitter and mistrustful of the state. Many Kamajors felt reintegration packages hardly compensated for their sacrifices and government administrators recognised that vocational training should have been better tailored to local market needs. In Uganda, soldiers’ theft of Arrow Boys’ demobilisation money undermined whatever legitimacy gains the army had earned through relatively successful cooperation against the LRA.

To be most effective disarmament, demobilisation and reintegration (DDR) initiatives should be gender-sensitive, taking into account the particular obstacles faced by female and male vigilantes.

After conflict, societies need to balance domestic and international calls for justice (by holding to account perpetrators of violence, including vigilantes) on the one hand, and calls for reconciliation to help communities confront the past and move on with their lives on the other.[fn]African countries emerging from conflict have sought, with varying degrees of success, to strike this balance. In Uganda, for instance, the 2007 Juba agreement, aimed at ending the Lord’s Resistance Army conflict, sought to prevent impunity for serious crimes and promote reconciliation by laying the groundwork for a mixture of formal and informal measures including a special division of the High Court to try atrocity crimes, local traditional justice mechanisms and amnesty provisions. Delays in implementing formal national justice and reconciliation schemes have led local communities and civil society to promote reconciliation at the local level. See Crisis Group Africa Report N°146, Northern Uganda: The Road to Peace, with or without Kony, 10 December 2008, pp. 8-11; and “Victims Fighting Impunity: Transitional Justice in the Great Lakes Region”, International Center for Transitional Justice, March 2017.Hide Footnote Key to striking a balance that reflects the priorities of affected communities is to ensure that victims have a role in designing national and local processes.[fn]Victims likely will have different opinions on justice and reconciliation that may change over time. Still, taking these opinions into account is critical to ensure legitimate and effective responses. See “Victims Fighting Impunity”, op. cit., pp. 9-17.Hide Footnote If widespread abuses have been committed on all sides, and if some combatants, including vigilantes, have been compelled to fight, affected communities might prioritise reconciliation over formal justice mechanisms. Justice and reconciliation initiatives optimally should be community-led and take account of cultural specificities.[fn]In South Sudan, communities held reconciliation ceremonies for Arrow Boys who had committed abuses in the same way as they did for former LRA members. Emilie Medeiros, “Back but not Home: Supporting the Reintegration of Former LRA Abductees into Civilian Life in Congo and South Sudan”, Conciliation Resources, August 2014, p. 5.Hide Footnote Former Kamajors who committed abuse struggled to gain social acceptance in their home areas; in the years that followed, community-level reconciliation became an essential part of their reintegration into civilian life.[fn]For many years after the war, local non-governmental organisations such as Fambul Tok (“Family Talk”) continued to help communities and former combatants reconcile perpetrators and victims (www.fambultok.org). The degree of abuse by former combatants was found to be the single greatest determinant of their acceptance or rejection by family and community members. See Jeremy Weinstein and Macartan Humphries, “Disentangling the Determinants of Successful Demobilization”, Center for Global Development, Working Paper No. 69, 2005, p. 23.Hide Footnote Donors can play a key role in providing international expertise and financial resources to help partner governments plan and implement sufficiently generous, locally-tailored disarmament, demobilisation and reintegration programs. To be most effective DDR initiatives should be gender-sensitive, taking into account the particular obstacles faced by female and male vigilantes, whether they have fought or played supporting roles, and the social stigma they may encounter as they assume family responsibilities or seek employment. Donors also could support civil society groups’ efforts to bring victims’ voices to the fore.

Governments should set realistic expectations to help mitigate the risk of alienating large numbers of unemployed former combatants.

Post-conflict management of groups that have fought on the government’s side is more challenging when they hope for rewards in the form of jobs in the security forces (as in northern Nigeria) or formalisation as an independent state-funded, entity (as in South Sudan). Governments should set realistic expectations to help mitigate the risk of alienating large numbers of unemployed former combatants.

Recruiting a significant proportion of former vigilantes into state security forces may be difficult because of their typically low education levels and large numbers.[fn]In Sierra Leone and Uganda, the armies’ educational requirements prevented many former vigilantes from joining. Crisis Group interviews, former Kamajor who joined the army, Freetown, 11 January 2017; Teso Arrow Boys’ administrative officer, Soroti, 29 January 2017.Hide Footnote But there are other roles they can play. Governments and their international backers should consider alternatives to deal with demobilised vigilante groups in a manner that minimises their discontent and, at the same time, makes the most of their local roots and, where applicable, legitimacy. For instance, former vigilantes might be retrained as unarmed community police units with the authority to gather information or even apprehend suspects. In either scenario, they would need adequate training, resources and oversight to take on these responsibilities.[fn]The viability of such initiatives depends greatly on local conditions. In Afghanistan, too little oversight and training meant that the Afghan Local Police – an experiment in semi-formal community policing supported by the U.S. – worsened security in many places. Crisis Group has argued for integrating the few effective units into the Afghan National Police and disbanding the rest. See Crisis Group Asia Report N°268, The Future of the Afghan Local Police, 4 June 2015.Hide Footnote

In most of Nigeria’s 36 states, where the largely federally-controlled security structures often fail to monitor or respond to grassroots insecurity, state governments have set up supplementary community police organisations or empowered community-based vigilantes.[fn]For instance, the Lagos Neighbourhood Safety Corps was launched in March 2017 with an initial 5,700 members; and the Niger state government legalised vigilante groups to assist the police, including by allowing them to arrest suspects even as it banned them from carrying weapons. The federal parliament is considering a bill to formally recognise the Vigilante Group of Nigeria (VGN), while the presidential adviser on the Niger Delta recently indicated plans to recruit 10,000 youths to guard petroleum pipelines. See, for example, “Ambode recruits 5,700 neighbourhood corps officials, deploys 4,554 vehicles”, Vanguard, 28 March 2017; “Niger govt okays vigilante groups for LGs”, Vanguard, 9 July 2013.Hide Footnote In the north east, civilians have a history of mistrust toward security forces, which they view as ignorant of local ways, arrogant, abusive and professionally incompetent. But CJTF members, by participating in efforts to counter Boko Haram, acted as a bridge between civilians and security forces, helping the state regain a measure of local legitimacy while protecting the local community. Giving former CJTF members a sense of purpose and responsibility in community policing roles in a close working relationship with state institutions could help prevent them from becoming a long-term security headache, and build on the positive outcomes of state-civilian security cooperation during the Boko Haram insurgency.[fn]See Crisis Group Report, Watchmen, op. cit., p. 21.Hide Footnote

IV. Conclusion

Relying on non-state armed actors to counter insurgencies might well be a necessary evil – but it ought to be a limited and finite one. The gravest dangers are posed when vigilantes pursue their own political-ethnic agenda; lack strong command and control structures, enabling battlefield commanders to promote their own interests; are largely unsupervised by either local or national authorities; or are ignored, unrecognised and cast aside once their military utility has expired. Support by an outside power against the wishes of the central state also increases the risk that vigilantes will fuel greater insecurity.

To limit the odds that vigilantes will turn from community protectors into insurgent forces, national leaders need to cooperate closely with local leaders and patrons to agree on a narrowly circumscribed mandate, geographic focus, and effective demobilisation, disarmament and reintegration. Under the best of circumstances, such an approach can do more than achieve short-term security gains. It also can help the central state forge closer ties to local communities, earning it the legitimacy needed to build longer-term peace.

Nairobi/Dakar/Brussels, 7 September 2017

Appendix A: Map of Sierra Leone

Appendix B: Map of Uganda

Appendix C: Map of South Sudan

Appendix D: Map of the Lake Chad Basin

Appendix E: Overview of the Four Cases of Vigilante Groups

Subscribe to Crisis Group’s Email Updates

Receive the best source of conflict analysis right in your inbox.