first to know

Robert Bain and Roseanne Burke

First to Know:Civilian-led early warning in armed conflict

Cover photo:South Sudanese soldier, Paloch, South Sudan, March 2014

© Punghi/Shutterstock

© Ceasefire Centre for Civilian Rights March 2021

This report has been produced with the financial assistance of the UK government.The contents of this report are the sole responsibility of the publishers and can underno circumstances be regarded as reflecting the position of the UK government.

Ceasefire Centre for Civilian Rights

The Ceasefire Centre for Civilian Rights is an international initiative to developcivilian-led monitoring of violations of international humanitarian law or humanrights, to pursue legal and political accountability for those responsible for suchviolations, and to develop the practice of civilian rights. The Ceasefire Centre forCivilian Rights is registered as a charity and a company limited by guaranteeunder English law; charity no: 1160083, company no: 9069133.

Community Empowerment for Progress Organisation

Community Empowerment for Progress Organisation is a South Sudanese NGObased in Juba with a country-wide network of activists that works to develop SouthSudanese society and civil society. CEPO’s activities include communityengagement, peacebuilding, conflict mitigation, democratic transformation,defending and strengthening human rights and the rule of law, developing viablelivelihoods and humanitarian work. www.cepo-southsudan.org

Acknowledgements

This report was reviewed by David Nyheim and Naomi Pendle, and copy-editedby Sophie Richmond. Any mistakes are the sole responsibility of the publishers.

Report designed by Staša Sukic.

Material from this publication may be reproduced for teaching or other non-commercial purposes, with appropriate attribution. No part of it may be reproducedin any form for commercial purposes without the prior express permission of thecopyright holders. Published March 2021. Printed in the UK on recycled paper.

Table of contents

Acronyms 4Introduction 5

The history of early warning systems 7

Piloting civilian-led early warning in South Sudan 23

Conclusion: The promise and challenge of civilian-led early warning systems 37

Endnotes 39

123

ACoH Agreement on the Cessation of Hostilities ARCSS Agreement on the Resolution of Conflict in South SudanAU African UnionCEPO Community Empowerment for Progress Organisation CEWS Continental Early Warning System CoHA Cessation of Hostilities Agreement CSO civil society organizationCTSAMM Ceasefire and Transitional Security Arrangements Monitoring Mechanism CTSAMVM Ceasefire and Transitional Security Arrangements Monitoring and Verification Mechanism DPKO Department of Peacekeeping Operations DPPA Department of Political and Peacebuilding Affairs

DRC Democratic Republic of the Congo ECOWARN ECOWAS Warning and Response NetworkECOWAS Economic Community of West African StatesEWS early warning system FCE Foundation for Co-ExistenceIGAD Intergovernmental Authority on Development JMAC joint mission analysis centre JMEC Joint Monitoring and Evaluation CommissionMVM Monitoring and Verification Mechanism NGO non-governmental organizationR-ARCSS Revitalised Agreement on the Resolution of Conflict in South SudanUN United NationsUNHCR UN Refugee AgencyUNMISS UN Mission in the Republic of South Sudan

Acronyms

Introduction

For decades it has been recognised that systematic early warning of armed

conflict has the potential to save lives. But in an era of widespread non-

international conflicts, which have had a devastating effect on civilian

populations, two questions have acquired a new urgency: who is doing the

warning and who is being warned?

The world’s newest state, South Sudan, descended into armedconflict in December 2013, a mere two years after it won itsindependence from Sudan. The scale of the violence and thesystematic targeting of civilians during hostilities and, inter-mittently, during ceasefires, have horrified the internationalcommunity. The international response to the war was swift,if not very effective. The first ceasefire was brokered just overa month after the outbreak of hostilities, but it did not proveto be durable or effective in stopping the violence.

The first ceasefire agreement established the first monitoringmechanism under the aegis of the eight-country East Africantrade bloc the Intergovernmental Authority on Development(IGAD), but this mechanism and its two successors have fo-cused on monitoring the ceasefires between the warring fac-tions and compliance with an agreement on the cessation ofhostilities that have repeatedly been breached. Civilians havecontinued to bear the brunt of the ongoing violence.

In contrast, the pilot early warning system (EWS) run bythe Community Empowerment for Progress Organisation(CEPO), a South Sudanese non-governmental organization(NGO), and the Ceasefire Centre for Civilian Rights is both

civilian-focused and civilian-led. Since it was launched inspring 2019, CEPO and Ceasefire have built networks to re-port actual and threatened incidents of violence againstcivilians, and successfully engaged key international deci-sion makers to further the protection of civilians from theviolence that continues to engulf South Sudan.

This report considers aspects of the role of civilians andcivil society organizations (CSOs) in early warning andsummarizes the lessons learned from the CEPO/Ceasefireearly warning project. The first section provides back-ground by detailing the development of early warning sys-tems for various forms of political violence – as opposed to,for example, early warning systems for natural disasters –from the 1960s until the present, tracing the developmentof the motivations and the conceptions of security on whichthey have been based.

The second section considers the pilot CEPO/Ceasefire EWSin South Sudan, considering the context in which it oper-ates, detailing the challenges it has faced and the opportu-nities it has realized, focusing on the lessons learned inbuilding and operating a civilian-led civilian-focused EWS,

6

and the challenges CEPO has faced in running it.It examines the difficult environment in whichCEPO operates, detailing the primacy that securityconcerns for staff and EWS monitors must be af-forded in the implementation of the project, andpolitical considerations that arise from operatingin such a sensitive situation. It also includes casestudies of how CEPO engaged several key groups– women’s activists and transport companies, set-ting out the challenges, frustrations and consider-ations encountered in expanding the network.

The report makes the case that civilian-led moni-toring systems should not be regarded as cheapalternatives to international expert-led teams. Itargues that these systems can be used to put thesafety of civilians at the centre of peace processesand bring civilian perspectives on violence, whichdo not privilege ‘political’ violence over thatwhich can be categorized as criminal or commu-nal, into political processes that may otherwise bedominated by negotiations between elites and se-curing ceasefires between warring factions.

First to Know: Civilian-led early warning in armed conflict

The history of earlywarning systems1

What are early warningsystems?Early warning systems collect and analyse data for theidentification of at-risk populations to inform decision mak-ers, who then may respond or intervene according to theinformation they receive. Early warning systems are usedfor a host of different purposes across different sectors, in-cluding climate change, natural disasters, pandemics, andarmed conflict. Conflict early warning systems identify keyrisk indicators for the outbreak of armed conflict. Conflictearly warning has been defined as:

‘a process that: (a) alerts decision makers to the potentialoutbreak, escalation and resurgence of violent conflict;and (b) promotes an understanding among decisionmakers of the nature and impacts of violent conflict.’ 1

Early warning systems have been described as a ‘risk man-agement tool’ that monitor and analyse different events

and variables of a local context to forecast whether thereis a risk of conflict.2 The different stages of early warningare:

• Collection of data (using specific indicators) • Analysis of the data (attaching meaning to indicators,

setting it in context, recognition of crisis development) • Formulation of best/worst scenarios and response

options • Communication to decision makers3

The definition of early warning is sufficiently broad thatmany different activities and approaches qualify as earlywarning. For example, an international NGO with one staffmember who monitors news reports and conducts regulartelephone conversations with people in conflict-affectedareas to produce reports would qualify as an early warningsystem, just as a multimillion dollar software system withproprietary algorithms and staff based across several coun-tries who conduct field research to produce alerts is alsoan early warning system.

Early warning is premised on the belief that violent conflict does not occur

spontaneously, and that there are key indicators or signs in the build-up to

conflict that, if properly identified, can alert monitors that there is a risk of

conflict escalation, outbreak or relapse. Importantly, it is also premised on the

belief that if the risk of violence is properly identified, the outbreak or escalation

of conflict can be prevented or, at the very least, its consequences mitigated.

8

It should be noted at the outset that all conflictearly warning systems face certain challengesthat set them apart from other early warning sys-tems, and impact on their effectiveness. First, un-like early warning systems which measurescientific phenomena, such as extreme weatherincidents, conflict early warning systems measuresocio-political factors and rely on predictinghuman behaviours, which cannot necessarily bemeasured by definitive indicators. Whereas sci-ence will determine how weather patterns de-velop, human behaviour can be affected by arange of contextual factors, including culture orhistory, which can mean that similar situations orincidents in different locations can result in radi-cally different outcomes. This also poses uniqueopportunities – hurricanes or diseases cannot benegotiated with, early warning systems merelyhelp to mitigate their effects, whereas conflictearly warning can theoretically prevent war fromoccurring at all.

Nevertheless, conflict early warning is more con-troversial than other forms of early warning. Dis-aster early warning will often act as an invitationto the international community to provide fund-ing and much-needed resources to avert disaster,whereas conflict early warning may result in re-sponses that threaten the internal political affairsof a state. The politicized nature of many factorsin conflict early warning can therefore make itcontroversial, which can be further exacerbatedwhere the early warning system is owned by anintergovernmental or political organization, suchas the United Nations (UN).

Generally speaking, most early warning systemsin operation today rely on software for data col-lection, input, and analysis against a set of indica-tors. However, specialized software may requirecertain technical skills, and this could exclude ac-tors without those skills or without access to therequisite technology.

Over decades, academic research has produced arange of conflict-related indicators by testing hy-potheses against existing data on previous armedconflicts, including identifying indicators for dif-ferent kinds of armed conflict. For example, an in-flux of arms to a country or an increase incirculation of arms will be a conflict indicator inall types of armed conflict, but poverty, large pop-

ulations and political instability have been iden-tified as key risk indicators for civil conflict specif-ically.4 However, standardized metrics of conflictmay lack context and thus fail to capture the nu-ances and risks of a particular situation on theground, while adding an air of unassailable au-thority that marginalized groups suffering fromthe conflict may find difficult to challenge.

Indicators can be quite broad – assessing state le-gitimacy as an indicator, for example, could in-volve analysing a range of considerations fromwhether there have been riots and protests in astate to whether the government is representativeof the population.5 Continuous monitoring of datacan be used to produce a baseline analysis of dif-ferent indicators. As data continues to be collectedand analysed, trends or fluctuations against theordinary baseline can be identified to indicatewhether there is an increased risk of conflict,which should then generate an alert – whether viasoftware or a person raising the alarm or both.Alerts or warnings can also be produced for sig-nificant events or incidents, such as a violent bor-der clash, which may indicate imminent risk ofconflict. Early warning systems can also measurecooperative or peaceful indicators, such as politi-cal agreements or legal reforms, which can helpto contextualize the broader situation in a state.

Once a report or an alert that indicates a risk ofarmed conflict is produced, early warning mustthen be followed by early response. The responseelement is recognized as one of the most challeng-ing aspects of early warning, which is affected by‘broader institutional, political, and contextual re-alities’.6 The warning and response elements ofearly warning systems have evolved significantlyover time, yet still face notable challenges in theirimplementation.

The fact that many civilians across the world con-tinue to face the horror of armed conflict andmass atrocities has led many to assert that earlywarning systems are not effective. However,given the complex array of factors that may deter-mine whether a country will descend into armedconflict, and how stakeholders respond at the pre-conflict stage, it is perhaps best not to view earlywarning systems as accurate predictors of allarmed conflicts – or mechanisms for reliably pre-venting them. Rather, they are a tool to raise the


9

alarm where key risk factors for conflict are pre-sent, which can allow for timely interventions toprevent violence arising or escalating.

Origins of earlywarning systemsEarly warning systems have their origins in sev-eral different fields, namely, military intelligence,conflict resolution and disaster early warning.

Military intelligence is arguably as old as war it-self, and generally involves monitoring and infor-mation gathering on political and militarydevelopments to determine potential military at-tacks and analyse the impact of potential conflictsupon a state’s or region’s interests. As technologyhas advanced, so has military intelligence; com-puterized systems linked to satellites, for exam-ple, can now provide timely information on themovement of arms and troops, or the deploymentof nuclear weapons.7

While early warning systems have some similar-ities to military intelligence, there are importantfactors that differentiate them from intelligencegathering.8 Early warning systems rely primarilyon open source information, such as newssources, social media and NGO reports,9 whereasintelligence tends to rely on clandestine informa-tion gathering. Importantly, early warning sys-tems take a global – and supposedly neutral –perspective with regard to early warning, as op-posed to an intelligence approach which analysesinformation and calculates responses through theinterests of a particular state or power.10 There-fore, although states have traditionally relied onearly warning for their own interests, early warn-ing systems which have emerged over the last 50to 60 years rely on the belief in an ‘internationalcommunity’ vested with powers to maintain in-ternational peace and security, and engage in con-flict resolution.

This shared global perception of an internationalcommunity largely developed after the creationof the United Nations in 1945 and the subsequentevolution of a global legal and diplomatic archi-tecture. Preventing armed conflict is one of thecentral aims of the UN Charter, which bestows

powers on the UN Security Council, the UN Secre-tary-General and the UN General Assembly to set-tle disputes, prevent violent conflict and maintainpeace internationally.11 A key tool in preventingconflict from escalating or arising is the use of‘preventive diplomacy’, a phrase coined by UNSecretary-General, Dag Hammarskjöld to describeefforts at the international level to prevent theCold War from escalating into a World War.12 Pre-ventive diplomacy entails a range of different ac-tivities, including soft measures such asfacilitating negotiation or mediation, coercivemeasures such as enacting sanctions or deployingpeacekeepers, and offering incentives such as in-vestment and financial aid. Today, preventivediplomacy is still a key element of conflict preven-tion and sometimes forms part of the response el-ement of early warning systems.

As the practice of preventive diplomacy was de-veloping in the political arena, conflict preventionalso came to be increasingly examined inacademia.13 In examining the root causes andshared characteristics of past conflicts, it was be-lieved that potential future conflicts could beidentified and prevented. Academics played sig-nificant roles in developing early warningmethodologies which formed the basis for thefirst conflict early warning systems. In the 1960s,the World Event Interaction Survey (WEIS) wascreated by Charles McClelland at the University ofSouthern California, which is sometimes referredto as the first conflict EWS.14 Data was collectedfrom New York Times articles on ‘events’ globallythat were either ‘cooperative’ or ‘conflictful’events. Examples of ‘conflictful’ events were riots,demonstrations, coups and assassinations. Analy-sis found that impending crises were indicated bythe intensity and range of different types ofevents that could be observed in some countries.15

At the same time, the idea of early warning sys-tems for humanitarian disasters was also increas-ingly developing. However, sophisticated disasterearly warning systems began to appear from the1980s onwards, while conflict early warning sys-tems remained largely similar to the WEIS modeluntil the 1990s. For example, the Famine EarlyWarning Systems Network (FEWS NET) was cre-ated in 1985 by USAID in response to famines inEast and West Africa. In addition to using satelliteimages to monitor the possibility of floods and


10

droughts, FEWS NET also deploys field monitorson the ground who collect and analyse data on theconditions in individual countries.16 The systemallows for targeted responses from states, donorsand international agencies in the form of aid, sup-ply chains and contingency planning.

Early application of conflict early warning also oc-curred to some extent through disaster earlywarning systems, because of the potential of con-flict to create humanitarian disaster. The UNRefugee Agency (UNHCR) and the former UN De-partment of Humanitarian Affairs (UNDHA), forexample, monitored population movements andrefugee flows to enable effective contingencyplanning, which also involved assessing risk ofconflict.17 The UNHCR High Alert List for Emer-gency Preparedness still includes risk indicatorsrelated to conflict.18

In 1992, building on the advances in the fields ofdisaster early warning systems, preventive diplo-macy and conflict prevention, UN Secretary-Gen-eral Boutros-Ghali advocated for the creation of aglobal UN conflict early warning system in his‘Agenda for Peace: Preventive Diplomacy, Peace-making and Peace-keeping’.19 While a comprehen-sive UN conflict early warning system becameensnared in the sensitivities of member states andnever materialized, the majority of conflict earlywarning systems emerged from the 1990s on-wards. This was partly driven by the failure of theinternational community to intervene in thegenocides that occurred in the mid-1990s inBosnia-Herzegovina and Rwanda, which providedrenewed impetus among the international com-munity to allocate funds and resources to pre-venting conflict and atrocities. The contexts andjustifications behind the development of thesevarious regional systems are investigated below.

At the international level, this led to the creationof the Office of the Special Adviser of the Secre-tary-General on the Prevention of Genocide in2004, and the adoption of the World Summit Out-come in 2005, which included the affirmation ofan international norm on the Responsibility toProtect (R2P)20 from genocide, war crimes, ethniccleansing and crimes against humanity.21 TheWorld Summit called for an expansion of UN ca-pabilities for early warning and assessment,22 andreaffirmed the principle that the international

community had not only the legal powers but alsothe moral duty to monitor situations of concernand intervene where necessary.

R2P is also happening alongside growing conver-sations about protection of civilians in the UN Se-curity Council, and growing humanitariancommitments to protection.

The adoption of R2P was also followed by the cre-ation of a Special Adviser on the Responsibility toProtect in 2008, and was accompanied by discus-sions about and commitments to the need to pro-tect civilians in the UN Security Council. The R2Pmandate has now been integrated into the Officeof the Special Adviser on Genocide, to create theUN Office on Genocide Prevention and the Re-sponsibility to Protect (OGPRtoP). The Special Ad-visers are mandated to provide an early warningfunction to the UN Secretary-General and theUNSC. OGPRtoP collects information, mostly fromUN sources, on political, human rights, humani-tarian, social and economic developments incountries worldwide to identify early warningsigns of the risk of atrocity crimes.23 Informationis analysed against the Framework of Analysis forAtrocity Crimes – the early warning methodolog-ical tool developed by the Special Advisers.24 Thetool outlines 14 different ‘risk factors’, with eachrisk factor listing several different indicatorswhich show that a risk factor is present.

Aside from the UN early warning system for atroc-ity crimes, a small number of other early warningsystems focus exclusively on mass atrocities, suchas Genocide Watch and the Sentinel Project.25

These systems draw heavily on academic theoryaround the root causes of genocide and identity-based violence, including Gregory Stanton’s tenstages of genocide model, which was originally de-veloped in the 1980s.26 Atrocity prevention andconflict prevention are separate but interlinkedfields which both developed in the latter half ofthe twentieth century, yet conflict early warningsystems far outnumber atrocity early warning sys-tems. This is perhaps because while atrocities mayoccur outside of armed conflict, conflict earlywarning systems should have robust indicatorsfor identity- or group-based violence, whichshould allow for atrocity early warning, andwould mean that atrocity-based systems would bea duplication of work and resources.


11First to Know: Civilian-led early warning in armed conflict

Originally called the State FailureTask Force, PITF was a projectdeveloped in 1994 that was fundedby the Central Intelligence Agencyof the USA. PITF was made up of a

Political InstabilityTask Force (PITF)

panel of scholars andmethodologists who aimed todesign and carry out a data-drivenstudy on the correlates of statefailure since the mid-1950s. Fromthis, the task force developed aglobal forecasting model to identifythe risk of ethnic war, revolutionarywar, adverse regime change, and

genocide.32 The project was run byacademics who were based indifferent universities across theUSA and relied on open accessdata, and at one point claimed tohave a predictive capacity ofbetween 80 and 90 per cent.33 Theprogramme was defunded by theUS government in 2020.34

Early warning systems have therefore been madepossible by global developments in politics, diplo-macy, civil society, academia and technology, anddevelopments in these areas continue to changethe way that early warning systems function andevolve. Systems that first emerged in the 1990sare markedly different from today’s early warn-ing systems, which has led to conflict early warn-ing systems being categorized into four different‘generations’.

Generations of conflictearly warningSince the 1990s, a plethora of different earlywarning systems emerged as the idea of conflictprevention and early warning increasinglygained traction internationally. Systems havebeen created by a variety of different governmen-tal, intergovernmental and non-governmental ac-tors at the local, regional and international level.In order to understand some of the key challengesand opportunities currently facing early warningand response, it is necessary to examine howearly warning systems have evolved, who the keystakeholders in these systems are perceived to be,and the benefits, disadvantages and criticisms ofdifferent models of early warning systems.

Conflict early warning systems have been catego-rized into four different generations.27 Categoriza-tion is not based solely on when systems appearedchronologically. Rather, the categorization of ‘gen-eration’ largely depends on: the data gathered –whether it is quantitative, qualitative or mixed;and who collects and analyses data, includingwhere the systems are based.28

First-generation systemsFirst-generation systems are predictive and de-signed to inform the decision making of internalclients, but do not usually have an in-built re-sponse process.29 Many first-generation systemsare therefore focused only on data gathering andanalysis, and do not carry out other functionsnow seen as crucial to early warning, such as for-mulating response options and engaging in advo-cacy.30 These systems are based in the globalNorth yet have a focus on monitoring countries inthe global South. They are generally quantitativein nature, and rely on a range of empirical datafrom news sources, as well as reports from NGOs,academia and intergovernmental organizations.31

In the case of PITF (see box) the internal clientwas the US government, which funded the projectfor policy-making purposes. Aside from publish-ing the results of the project, there was no re-sponse element involved in PITF: any responsewould depend on how the US government choseto use the analysis provided, which was inevitablyaffected by the interests and priorities of the USas a state. First-generation systems are generallyaimed at ‘track one’ actors: states and intergov-ernmental organizations such as the UN,35 buthave little to no effective communication proce-dure for their findings.

One benefit of initiatives such as PITF is that thesesystems can include more political factors in riskassessment models than those included in the riskassessment models of intergovernmental or re-gional organizations due to political sensitivities,(e.g. regime type, state-led discrimination andlength that a leader has been in office).36 How-ever, the fact that any response is expected to beundertaken by track one actors means that the

12

analysis provided by any first-generation systemwill still be constrained by these political consid-erations.

In addition to the gap between warning and re-sponse of first-generation systems, their relianceon quantitative data and lack of input from actorson the ground raises serious problems over accu-racy, as statistical approaches cannot fully ac-count for case- and context-specific information.37

Importantly, first-generation systems lack the ex-pertise and input of local actors, who are not onlythe most affected by armed conflict but are alsocrucial for understanding local contexts and cre-ating tailored responses. While it is accepted thatfirst-generation systems can be useful for moni-toring and recognizing trends in armed conflict,38

they are no longer seen as effective early warningsystems.

Second-generation systemsSecond-generation early warning systems have aresponse element, usually in the form of proposalsor recommendations aimed at track one actors.39

These systems also conduct monitoring in conflict-affected countries and regions, but the analysis ofdata is usually carried out in the global Northwhere many of the systems are headquartered.Many of these systems rely on qualitative data, al-though some rely on a mixed methodology, usingboth quantitative and qualitative approaches.

Although ICG (see box) does engage with local ac-tors, its advocacy efforts are largely aimed attrack one actors, particularly the UN and the Eu-ropean Union (EU).41 This means that any re-sponse is constrained by the political sensitivitiesand interests of the international community (orlack thereof) to intervene. Many of ICG’s fieldmonitors are not locals and are instead from the

global North. They either travel to conflict-af-fected countries on dedicated trips and/or arebased in or near conflict zones. While there is anargument that international NGO workers whoare not from the local country or region may beseen as more impartial or independent, the actualor perceived neutrality of international actors isdebatable.

There are numerous criticisms of second-genera-tion models. They tend to rely on qualitative data,which it has been argued may be too subjective.While qualitative data can offer richer and morenuanced accounts of what is occurring on theground than empirical analysis only, the volumeof data can also present problems.42 Further criti-cisms of second-generation models are also rele-vant to first-generation systems. First, the fact thatthose who are carrying out analysis and advocacyefforts are based in the global North, as well as thefocus on track one actors as the agents of re-sponse, limits the potential effectiveness of re-sponses. This is because ‘early warning agenciesin the West have little influence over the institu-tions that could execute preventive diplomacy inactual conflict’, and ‘the distance between the pos-sible conflict areas and the West lessens the inter-national community’s motivation to intervene’.43

Those who are based in the field lack the abilityto perform response functions on the ground dueto the ‘top-down’ approach of second-generationsystems, and the fact that field monitors are oftennot locals with relevant local knowledge.

Furthermore, first- and second-generation sys-tems ignore micro-level tensions and indicators ofconflict because of their focus on states. Thismeans that tensions or potential outbreaks of vi-olence between different communities that don’tinvolve state actors may not be adequately cap-


ICG was established in 1995 inresponse to the failure of theinternational community to act toprevent mass atrocities and genocide

International Crisis Group (ICG)40

in Bosnia, Rwanda and Somalia. It isheadquartered in Brussels and hasstaff based in or nearby conflict-affected countries who conduct fieldresearch and engagement. This islargely done through conductinginterviews with a range of differentstakeholders, including local leaders

and government officials. Evidencegathered is used to inform a globalconflict tracker and to produce amonthly early warning bulletin. ICGundertakes high-level advocacybased on its findings, and states onits website that it is ‘influential inWestern power centres’.


The most oft-cited example of athird-generation early warningsystems is the FCE initiative in SriLanka, which started in 2002. Acitizen-based model of earlywarning was designed, as well ascomputer software called FCEWARNwhich specifically monitoredconflicts at the micro-level, whichwas used alongside GeographicInformation Systems. Data wasprovided by field monitors operatingin the conflict zone who were

Foundation for Co-Existence (FCE)48

members of local communities, anda network of committees at thegrassroots level –made up ofrepresentatives from a broad rangeof different demographic groups –also provided information. Membersof these committees were trainedand mobilized to monitor andidentify the peace and conflictindicators at local levels.

There was also a parallel responsemechanism through which ‘multitrack diplomacy’ was used by localstakeholders in order to interveneand prevent conflict. Once asituation had been flagged by the

software, field monitors,committees and informationanalysts analysed the informationthey had and developed aprognosis for the issue concerned,then identified and classified thestakeholders and potentialinterveners. After establishing howimminent the threat was, links withlocal stakeholders and intervenerswere established and a responsewas initiated. Innovatively, thesystem also had an SMS function,which alerted key local actors whohad the ability to de-escalateviolence to the fact that there wasan imminent risk in their area.

tured. Moreover, when it comes to resolving thesemicro-level issues, the international communitywould not be the appropriate responder.44

Finally, first- and second-generation systems raiseimportant questions about who early warningsystems are designed to warn. The fact that localcommunities are not even recipients of earlywarning information in these systems is deeplyproblematic. While the international communityundoubtedly has a role to play in early warningand response, the development of third- andfourth-generation systems represents a shift inthe way early warning systems are designed, andacknowledges the significant role that local com-munities have to play in conflict early warningand response.

Third-generation systemsThird-generation systems are entirely based inconflict-affected countries and regions. Field mon-itoring roles are assigned to those who live in con-flict-affected areas, and they also act as firstresponders to early warning indicators. These sys-tems therefore combine early warning and earlyresponse, and use mixed methodologies to collectand analyse data. Third-generation systems in-volve track one and track two actors, have a focuson micro-level conflict indicators and resolutions,and sometimes have mechanisms for includinglocal community leaders in responses.45 These sys-

tems are often referred to as ‘people-centred sys-tems’,46 or systems ‘of citizens, by citizens and forcitizens’.47

Third-generation systems are characterized bytheir local ownership. While the gathering of dataand use of potentially problematic computerizedsystems to produce early warning is similar tosecond-generation systems, the response elementof third-generation systems is markedly different.FCE has defined early warning as sending ‘theright information at the right time to the rightpeople to take timely action for prevention of con-flicts’.49 In third-generation systems, ‘the rightpeople’ are not always track one actors, but localactors who are able to take timely action.

However, third-generation systems do raise anumber of issues. The fact that they are micro-level and reactive rather than pro-active meansthat in practice they are rarely able to addressstructural and root causes of violence.50 Whilelocal stakeholders may be able to react quickly tocalm tensions in the short term, the underlyingcauses of those tensions will continue to exist un-less macro-level state action, or even action at theinternational level, is taken in order to addressroot causes of violence or disagreement. Finally,while these systems are locally owned, they arestill not accessible to the public, and early warn-ing is targeted at specific stakeholders who it is

14

believed have the ability to prevent conflict,rather than notifying communities at risk.

Fourth-generation systemsLike third-generation systems, fourth-generationsystems are people-centred systems that are en-tirely based in the conflict region. However, theyhave less centralized structures and rely directlyon civilians and civil society to provide data.Whereas second- and third-generation systemshave designated field monitors and formal report-ing structures, and usually use proprietary soft-ware, fourth-generation systems are ‘horizontal’,facilitating reporting and warning by and withincommunities51 These systems use open source soft-ware which allows ordinary citizens to reportevents, and aim to empower individuals and com-munities to take action to reduce the potential riskthat latent conflict poses directly to communities.52

There are numerous benefits to fourth-generationsystems and the use of civil society-led earlywarning systems. They allow for real-time report-ing and a plurality of different perspectives thatis more inclusive than other generations of earlywarning systems, particularly because many sys-tems allow for reports and events to be docu-mented in a variety of different languages. Thefact that they are participatory removes the timelapse of using intermediaries to report or react toconflict warning signs, and these systems can ef-fectively leverage the expertise of local civil soci-

ety actors who not only have in-depth knowledgeof local contexts, but are also often seen as legiti-mate and trusted within their communities.58 Fur-thermore, this participatory approach, and theopen nature of the information reported, meansthat fourth-generation systems truly function towarn those who are most vulnerable. Aside fromcivil society or other actors who are able to use in-formation in order to try to resolve conflict, peo-ple living in conflict areas are able to viewinformation about the scale, frequency and sever-ity of events occurring around them and to makedecisions about how best to protect themselves,their family, their livelihoods and their property.

As noted in the section on third-generation sys-tems, removing state or track one actors alto-gether is not desirable, as sustainable solutionswill usually need the support of state or interna-tional actors. However, fourth-generation systemshave allowed civil society to use information gath-ered through these platforms in order to conductadvocacy at the national and international level.Furthermore, some CSOs are able to feed theirearly warning data into other state-level or re-gional systems that primarily inform govern-ments or intergovernmental organizations.

Rather than there being one ideal model or gen-eration of early warning, there is a general con-sensus that all four generations of early warning


Ushahidi (which means testimony inSwahili), was created in 2008 to usecrowdsourcing in order to mapreports of post-election violence inKenya. Users submitted reports bytext message, smartphoneapplication, Twitter, email, or via thewebsite, specifying the time, locationand type of abuse.53 They could alsoinclude pictures, video evidence, orlinks to media stories to corroboratetheir reports. These reports weregeo-tagged and plotted on a map,producing a live map of the crisis.However, rather than warning of

Ushahidi impending violence, the systeminitially functioned to raise an alertwhen a crisis or violence was alreadyunder way.54 Over time, it hasdeveloped ‘from a simpleWordpress blog with dots on a mapinto an entire ecosystem of softwareand tools built to facilitate the workdone by human rights advocates,journalists, election monitors andthose responding to disaster andcrisis’.55 In addition to its reportingfunctions, the software now has arange of different tools andfunctions, including disseminatingsurveys, generating reports andanalysing open source data.

Ushahidi has developed a businessmodel whereby it provides its openplatforms to local and internationalNGOs to serve as an early warningand crisis mapping system. Over20,000 Ushahidi maps have beenlaunched in more than 140countries.56 For example, in 2015, itworked with local NGOs in Nigeriain the run-up to elections in orderto assess the possibility of violence.Software was developed whichassessed Twitter and news sourcesfor indicators of potential violence,and potential ‘hotspots’ wereflagged to local NGOs who carriedout election monitoring.57

15

systems can contribute to an ‘ecosystem ap-proach’ to conflict early warning and rapid re-sponse.59

The UN systems: gaps inwarning and response There have been several attempts to institute acomprehensive conflict early warning system atthe UN.60 The Office for Research and the Collec-tion of Information (ORCI) had an early warningmandate during its operation from 1987 to 1992.61

After ORCI, the Department for Humanitarian Af-fairs (DHA) established the Humanitarian EarlyWarning System (HEWS), which collected quanti-tative information on a range of countries of con-cern.62 However, the system reportedly was notable to produce a single early warning of armedconflict, was overly reliant on quantitative data,63

and was subsequently abandoned.64 It has been as-serted that the UN’s failure to establish a compre-hensive system is due to the fragmented nature ofearly warning carried out by different UN agen-cies and the resistance of member states.65

Early warning is carried out by seven UN bodiesand one ad hoc initiative; the Department of Polit-ical and Peacebuilding Affairs (DPPA), UN Devel-opment Programme (UNDP), Department ofPeacekeeping Operations (DPKO), Office for theCoordination of Humanitarian Affairs (OCHA),World Food Programme (WFP), Office for the HighCommissioner for Human Rights (OHCHR), Officeof the Special Adviser of the Secretary-General onthe Prevention of Genocide (OSAPG) and theGlobal Pulse (formerly known as the Global Im-pact and Vulnerability Alert System, or GIVAS).66

Several of these agencies incorporate conflict earlywarning within disaster or other humanitarianearly warning mandates, whereas some are moreconflict-focused: these are outlined below.

OCHA: Established in 1991 and formerly known asthe Department of Humanitarian Affairs, OCHA’smandate includes promoting preparedness andprevention for disasters and emergencies. It has anEarly Warning and Contingency Planning (EWCP)section which monitors social, economic, politicaland environmental indicators to assess risks, eval-uate trends and produce early warning products

in the form of ‘snapshot’ assessments and, if nec-essary, more in-depth analyses of the human secu-rity sectors in a given country. OCHA also chairsthe Inter-Agency Standing Committee (IASC),which coordinates humanitarian efforts among 17UN and non-UN partners under the leadership ofthe Emergency Relief Coordinator. OCHA preparesthe quarterly humanitarian-focused Early Warning– Early Action report, which represents the viewsof the IASC, including conflict early warning trendsand potential interventions. The report’s findingsare also referred to emergency directors for oper-ationalizing an emergency response.

UNDPPA: Originally established in 1992 as the De-partment for Political Affairs, DPPA is mandatedto prevent and resolve deadly conflict around theworld. In addition to managing political missions,DPPA produces analytical reports and briefingnotes on potential crises to the Under Secretary-General for Political Affairs (USGPA). The USGPAsits on the Secretary-General’s Policy Committeeand can raise potential crises with the Secretary-General to share with the Security Council. DPPAalso houses the Mediation Support Unit (MSU),which was established in 2006 and provides advi-sory, financial and logistical support to peace pro-cesses, as well as maintaining a roster ofmediation experts who can be deployed to engagein mediation.67

DPKO was established in 1992, with a mandate toplan, manage and deploy peacekeepers when au-thorized by the Security Council. It maintains a 24-hour Situation Centre which has two earlywarning components, the Operations Room (OR)and the Research and Liaison Unit (RLU). The ORreceives information from the field, and monitorsnews sources to produce incident reports and amonitoring update to senior leadership threetimes weekly. The RLU takes a longer-term ap-proach to early warning; it collates data and pro-duces early warning reports of political, militaryand security trends that affect ongoing or potentialpeacekeeping operations for senior decision mak-ers. In 2006, DPKO also created joint operationscentres (JOCs) and joint mission analysis centres(JMACs) which both work with the Situation Cen-tre. JOCs collect situation updates and provideshort-term analysis to the head of mission and thesenior management team, whereas JMACs providelonger-term analysis to support decision making.


16

Developing early warning mechanisms falls underDPKO’s protection of civilians mandate,68 andsome peacekeeping missions are tasked with cre-ating a mission-specific early warning system. Forexample, UN Security Council Resolution 1996 es-tablished the United Nations Mission in the Repub-lic of South Sudan (UNMISS), and mandatedUNMISS to establish and implement ‘a mission-wide early warning capacity, with an integratedapproach to information gathering, monitoring,verification, early warning and dissemination,and follow-up mechanisms’.69 UN peacekeepingmissions in South Sudan, the Democratic Republicof the Congo (DRC), and the Central African Repub-lic (CAR), in particular, have established earlywarning mechanisms and networks for the pro-tection of civilians.70

Conflict contexts such as the DRC and South Sudanpose complex challenges because they feature anarray of state and non-state actors, as well as a va-riety of factors which drive conflict, including eth-nic, political, and economic tensions. Unlike otherearly warning systems, which take a global ap-proach to examining which states are most at riskof conflict, early warning systems needed inpeacekeeping missions are focused on forecastingthe areas at risk of violence within one state. Thismeans that some of the more traditional metricsused to examine the risk of conflict at the statelevel, such as the fragility of state institutions orcorruption, will not be as valuable. Instead, micro-level data is essential to understand local dynam-ics and potential threats to peace, which requiresaccess to a wide range of different communities orgroups living within the state in order to collectdata.

The UN mission in DRC (MONUSCO) responded tofailures to implement early warning and responseby creating a range of new tools to facilitate earlywarning, such as Community Liaison Assistantsand Community Alert Networks, that have nowbeen replicated in other missions.71 Community Li-aison Assistants are national staff who serve as aninterface between the mission and the local popu-lation, as well as with local authorities, and theyplay a key role in community engagement.72 Theycollect information on security threats and alertpeacekeepers to enable quick responses.73 Com-munity Alert Networks enable communitieswhere MONUSCO is present to alert MONUSCO to

violent incidents or threats to local communities,usually through equipping a designated commu-nity member with a phone or radio.74 This infor-mation is then transmitted by the mission to itsJOC and JMAC, who analyse reported informationagainst conflict indicators to assess the risk of con-flict and designate areas as red, yellow or green.75

However, significant challenges still remain in en-suring that effective early warning is in place. Thefact that much of the data gathered relies on inci-dent reporting means that often violence has al-ready occurred (65 per cent of reports in 2013),making the system more reactive than predictive.Lack of mobile network coverage in rural areascan also pose problems for reporting, and areasthat are not covered by the Community Alert Net-work are effectively excluded.76 Furthermore,even where communities are covered by the net-work, where they are located at a significant dis-tance from MONUSCO forces, the mission hasinadequate time to mobilize resources to respond.As MONSUCO has reduced its military footprint,this has led to an increased reliance on the Com-munity Alert Network To work with local actorsand authorities to resolve problems.77

Access therefore remains a key issue, and one thatis unlikely ever to be fully overcome by systemswhich rely on monitoring or data collection by ex-ternal rather than local actors. There have beeninstances where UN agencies and peacekeepershave been denied access to conflict-affected re-gions, such as in South Sudan,78 demonstrating thelimitations of relying on international actors to im-plement early warning and response.

Responding to early warningHistorically, criticism of the UN’s early warningcapacity has centred around a failure to respondto early warning signs, as occurred during thebuild-up to the Rwandan genocide when warn-ings by the DPKO, Special Procedures mandateholders and member states were not acted upon.79

More recently, UN systems have faced criticismbecause of a failure to communicate across agen-cies and systems that are all operating in parallel,which has led to inadequate responses to earlywarning signs.

In 2006, the UN International Strategy for DisasterReduction (UN/ISDR) produced guidance on devel-


17

oping early warning systems. It outlined four keyelements that must be present in order to develop‘people-centred’ early warning systems: (1) RiskKnowledge – systematic collection and analysis ofdata; (2) Monitoring and Warning Service – a reli-able forecasting and warning system; (3) Dissemi-nation and Communication – warnings must reachthose at risk; and (4) Response Capability – buildnational and community response capabilities.80

UN country teams usually combine all of the UNagencies and programmes operating on theground in a country, and a Resident Coordinatormanages the coordination of the work of all theagencies,81 however UN agencies can work inde-pendently in practice and a degree of competitionbetween them is not unknown. Further, as manyof these agencies work are running humanitarianaid programmes, capacity for political analysisand diplomacy is limited and these aspects do notalways directly inform the work of missions on theground. Parallel to the of-country teams, theUNDPPA based in New York conducts analysis andearly warning and is principally focused on polit-ical analysis and diplomacy. While the work ofcountry teams and the Secretariat should there-fore be complementary, in the past, there has beena failure to communicate across different agenciesand strands of work.

In 2012, for example, an internal review of theUN’s work in Sri Lanka during the final years ofthe civil war found a systematic failure of both theUN country team and the UN Secretariat to prop-erly integrate their work on early warning and re-sponse.82 The review found that there was no jointanalysis of the threats to civilian populations, anda failure to communicate across UN agenciesmeant that responses to perceived threats wereweak or non-existent.83 The review’s findings con-tributed to the introduction of the Human Rightsup Front (HRuF) initiative which was launched byUN Secretary-General Ban-Ki Moon in 2013.84

One of the main objectives of HRuF was to im-prove early warning mechanisms and capacitywithin the UN to prevent conflict and serious vio-lations of human rights and humanitarian law.85

However, the failure of the UN to act to prevent vi-olence and atrocities in Myanmar suggest thatHRuF has not been able to remedy the UN’s his-toric problems of a lack of coordination and re-

sponse among agencies and failure to respond toclear warning signs:

‘while the UN’s risk analysis tools effectivelyidentified Myanmar as a crisis necessitating ahuman rights-oriented approach … senior UNleadership failed to resolve bureaucraticinfighting, set a common strategy, or establishconsistent messaging. This contributed toconfusion and paralysis in the face of a rapidlydeteriorating human rights and humanitariansituation.’ 86

The differing roles of both condemning and coop-erating with states that respective UN agenciesplay can pose complex problems for effective earlywarning. Furthermore, as an intergovernmentalorganization, UN decision making will always beinformed and influenced by global politics, andmember states have been wary of the conse-quences of early warning systems. It has been as-serted that the resistance of member states toearly warning is due to three reasons: ‘(1) a gen-eral aversion to being monitored by any outsideorganization for activities that occur within theirsovereign territory; (2) the risk that predictionswill embolden conflicting parties and become self-fulfilling prophecies; and (3) the risk of adverseeconomic consequences of “watch list’” designa-tion.’87 Further, the UN system is state-centric andthis has led certain actors to view conflicts andpossible solutions to those conflicts in a particularand often overly reductive manner. The constraints on early warning within the UNand the failure of the UN to effectively interveneraises questions not only over the warning and re-sponse capacities of UN early warning systems,but it also raises more fundamental questions overwho the principal actors in early warning shouldbe and who systems are designed to warn.

While the guidance may have had natural disasterearly warning in mind, all of these elements arealso essential for conflict early warning. Yet noneof the UN systems examined in this section couldbe described as ‘people-centred systems’. The UN’svertical approach to conflict early warning meansthat those most at risk are not those who receivewarning alerts, nor are they part of the responseelement. The different approach to conflict earlywarning as opposed to natural disaster demon-strates the consequences of the controversial na-


18

ture of the former. While the UN may lament thefailure of its systems to prevent conflict, the failureof those systems to communicate warnings tothose at risk has the most serious consequencesfor civilians living in conflict-affected areas.

Mapping earlywarning systems:challenges andopportunities

Since early warning systems first emerged a fewdecades ago, many systems have come and gone.88

In order to understand some of the present-daychallenges facing early warning, it is necessary toexamine the different systems that currently existat the international and regional level, and howthese systems coexist and interact, before lookingat national and local early warning in SouthSudan.

Regional systemsMany early warning systems are rooted in the be-lief that international actors have a responsibilityto act as protectors where they are alerted to therisk or outbreak of conflict, and that they will actupon early warning to prevent conflict and savecivilian populations.89 This belief has failed to beborne out in reality. The failure of internationalorganizations and Western-centric early warningsystems to prevent conflict has led to an increasingrecognition of the role that local and/or regionalinitiatives can and should be playing in earlywarning systems and conflict resolution. Therehas been increased support for regional earlywarning initiatives, the African Union’s Continen-tal Early Warning System (CEWS), for example,benefits from EU funding as well as collaborationwith the UN.

The African continent is the most covered regionglobally in terms of early warning systems, andboasts some of the most advanced regional earlywarning systems.90 Large strides have been madeover the past decade in operationalizing earlywarning systems across Africa, yet significantchallenges remain, as evidenced by the continuedstate of armed conflict that many citizens con-

tinue to find themselves in. Some of the most com-mon challenges that the early warning systems ofmultilateral organizations face are ‘(a) weak earlywarnings; (b) immature response mechanismsand instruments; and (c) personal, institutionaland political shortfalls’.91 While these shortcom-ings were identified in 2009, an examination ofthe regional systems in Africa suggest that theseareas remain key challenges for early warning adecade later.

CEWSThe African Union has a primary role for peaceand security on the continent, under the UN,which has responsibility for global peace and se-curity. The Constitutive Act of the African Union(AU) came into force in 2002, the AU succeedingthe Organisation of African Unity (OAU). This tran-sition marked a departure from the OAU’s found-ing principle of non-interventionism92 as the AU’sConstitutive Act gives the organization power ‘tointervene in a Member State pursuant to a deci-sion of the Assembly in respect of grave circum-stances, namely: war crimes, genocide and crimesagainst humanity’.93

The Peace and Security Council (PSC) is the deci-sion-making body of the AU, and it is mandated toprovide early warning to facilitate responses toconflict situations in Africa, and the African UnionCommission is responsible for carrying out PSCdecisions.94 The Protocol relating to the Establish-ment of the PSC provides for the establishment ofa Continental Early Warning System (CEWS), inorder to facilitate the anticipation and preventionof conflicts in Africa.95 The PSC and CEWS are twoof the five core pillars of the African Peace and Se-curity Architecture (APSA), along with the Panelof the Wise (PoW) (which undertakes preventivediplomacy), the African Standby Force (which in-cludes military, civilian and police elements andcarries out peace support operations), and thePeace Fund (a financial instrument to supportAPSA).

Through CEWS, the AU Commission is also man-dated to ‘collaborate with the United Nations, itsagencies, other relevant international organiza-tions, research centers, academic institutionsand NGOs, to facilitate the effective functioningof the Early Warning System’.96 The system con-sists of:


19

‘an observation and monitoring unit, to beknown as “the Situation Room”, which is locatedat the Conflict Prevention and Early WarningDivision of the African Union and is responsiblefor data collection and analysis; andthe observation and monitoring units of theRegional Mechanisms for Conflict Prevention,Management and Resolution, which are to belinked directly to the Situation Room.’ 97

Data is gathered from a variety of differentsources, including the media, academia, think -tanks, as well as data from AU sources (the Com-mission, AU Field Missions and Liaison Offices).98

CEWS also relies on data from sub-regional earlywarning systems that are run by Regional Eco-nomic Communities (RECs) across Africa includingthe Economic Community of West African States(ECOWAS), the Intergovernmental Authority onDevelopment (IGAD) and the Southern African De-velopment Community (SADC). All of these RECshave established their own early warning systems,many predating the CEWS. CEWS has developed aninternet portal to connect it with the early warningsystems of the RECs, however, the ECOWAS systemwas the only one fully connected as of 2017.99

CEWS uses software which can generate a numberof different outputs, including country reports,daily highlights, weekly summary reports andSMS alerts for the AU chairpersons and membersof the PSC and PoW. CEWS also produces policyrecommendations and possible response optionsto potential conflict situations. The CEWS method-ology outlines three clusters under which data willbe collected: ‘1. Context and Structural Informa-tion on Countries and Regions; 2. Actor AttributeInformation on key Individuals and Groups; and3. Information on Behaviours and Events as theyevolve over time.’100

CEWS conducts baseline analysis on political, eco-nomic, social, military and humanitarian indica-tors. As well as developing a continent-wide coreof shared indicators, the CEWS Handbook notesthat it is important also to develop country- andregion-specific indicators to monitor the uniqueaspects of conflicts in particular areas.101

CEWS relies on the effectiveness of the early warn-ing systems of the RECs, as well as the response el-ement undertaken by relevant AU actors. The

interoperability of the different early warning sys-tems is somewhat complicated given that the earlywarning systems of the RECs predate CEWS, andhave their own legal mandates, strategies andaims. While the AU envisages CEWS as the overar-ching system, the role of the RECs in both warningand response remains crucial, and the effective-ness of the different early warning systems of theRECs will impact on the effectiveness of CEWS.

Significant challenges to the operation of CEWS ac-knowledged by the AU are lack of funding andhuman resources, as well as lack of ICT infrastruc-ture to make the system operational to its full ca-pacity. Just as decisions by UN bodies remainsubject to some level of political influence, so toodo the decisions of the AU and other regional bod-ies: ‘interventions by the AU and/or regional orga-nizations in recent conflict situations shows thatthe individual member states continue to set clear“red lines”‘.102 This means that there has been alack of political commitment from member statesto use early warning information and put emerg-ing conflicts on the agenda.103

ECOWARNOriginally an economic organization, ECOWAS104

established its Warning and Response Network(ECOWARN) in its Protocol on Democracy andGood Governance in 2001, in response to the threatthat insecurity and conflict posed to economic sta-bility in the region.105 ECOWARN became opera-tional in 2003, it consists of an Observation andMonitoring Centre (OMC) headquartered in Abuja,Nigeria, with four sub-regional offices operating asmonitoring zones based in Banjul, Ouagadougou,Monrovia, and Cotonou (in The Gambia, BurkinaFaso, Liberia and Benin respectively).106

ECOWARN is not only a conflict early warning sys-tem, it tracks six different ‘sectors’, including se-curity, crime, health and the environment.107

ECOWARN uses an open source analytical frame-work with over 150 indicators across the sectorsto analyse vulnerabilities within member states. Itmakes use of ‘a network of field monitors, humansecurity indicators and state of art Geographic In-formation System tools providing real-time infor-mation and data’.108

ECOWARN also operates in partnership with theWest African Network for Peacebuilding (WANEP),


20

an umbrella body for CSOs operating across the re-gion to collect data to inform ECOWARN. WANEPcollects data ‘on human security issues, most no-tably human rights and democracy, food short-ages, unemployment, arms flows andcivil–military relations and droughts and flood-ing’, and prepares reports for the OMC.109

ECOWARN uses field monitors in each monitoringzone who provide constant updates of informa-tion, which is fed into a web-based visualizationtool, as well as to inform ‘daily highlights, situationreports, monthly country policy briefs, incident re-ports, security reports, and thematic reports’.110

ECOWARN also produces a Country Risk and Vul-nerability Assessment that allows member statesto identify patterns and trends in countries and re-gions. Several national initiatives in ECOWAScountries have also been developed which are de-signed to complement and feed into the ECOWARNsystem. For instance, Ghana, where WANEP has itsheadquarters, has launched a national early warn-ing system called ‘Ghanawarn’.111

ECOWAS has taken a militaristic approach to con-flict prevention. It has a Mediation and SecurityCouncil (MSC) which decides through a majorityvote on military interventions in member statesagainst the will of target countries, that is, ‘againstthe will of target countries in cases of, among oth-ers, violation of human rights, the rule of law, ordemocratic principles’.112 ECOWAS has used thesepowers to intervene militarily in Guinea-Bissau,Burkina Faso and the Gambia, on account of coupsand contested election results.

However, there remain many protracted conflictsin the region in which ECOWAS has not inter-vened. Although ECOWAS is mandated to inter-vene ‘in internal armed conflicts that havesecurity implications for countries in the sub-re-gion’,113 it has often not been invited to intervenein purely local conflicts without obvious regionalsecurity implications, such as the Niger Delta.

Further, questions have been raised over the abil-ity of early warning systems to tackle the newforms of conflict that have emerged over the pasttwo decades, especially those based on standard-ized indices that may be blind to the contextualfactors that drive human behaviour. The emer-gence of extremism, terrorism and criminalized

conflict poses problems for early warning systemsthat are grievance-based: for example, monitor in-dicators related to a population’s relationship withthe government.114 This is borne out by the failureof ECOWARN to predict or prevent armed terroristactivities in Mali and Nigeria in 2015.

Many conflicts, particularly in the African conti-nent, are also hybrid in nature. For example, theremay be several different conflicts that occur simul-taneously, or many different issues that drive con-flict, including criminality, extremism or climatechange.115 This also poses complex problems forthe response element of early warning systems –preventive diplomacy or mediation between op-posing political groups will require different ap-proaches, for example, than settling disputesbetween armed criminal or extremist groups overresources which neither have a legal claim to.116

On a more operational level, ECOWAS functions asa second-generation system which is highly verti-cal in nature and is bureaucratic. Major GeneralO.B. Akwa, Commandant of the Kofi Annan Peace-keeping Training in Ghana, stated that ‘the transi-tion from ECOWAS of States to ECOWAS of peoplehas not taken full effect yet’,117 noting that officialsare not always working in the interests of citizensdespite having the tools to serve them.118 A WANEPreport stated that: ‘Government should improve[the] civil–military relationship to foster trust andconfidence between security agents and the localcommunities for improved early warning alerts,information sharing and collaboration in fightinginsurgency at all levels.’119

Despite relying heavily on CSOs for data collection,ECOWARN is not a people-centred system, andCSOs are not recipients of early warning informa-tion. Warning alerts are provided to memberstates and relevant ECOWAS actors. AlthoughECOWARN faces significant challenges, includinga warning–response gap, and questions over theability of its systems to flag risk factors for themany different types of conflict present in the re-gion, ECOWARN is generally seen as the most suc-cessful regional early warning system in Africa.

CEWARNThe Conflict Early Warning and Response Mecha-nism (CEWARN) is the EWS created in 2002 by theIGAD, which is made up of eight states in the Horn


21

of Africa.120 It was established under IGAD’s Peaceand Security Division under the CEWARN Proto-col. CEWARN is ‘an interwoven network of govern-mental and non-governmental organisations’,121

and is generally regarded as the earliest third-gen-eration early warning system. In the initial yearsof its operation, it was only mandated to monitorpastoralist-associated conflicts, and the systemcovers three geographical ‘clusters’: (a) theKaramoja Cluster (cross-border areas of Ethiopia,Kenya, Sudan and Uganda); (b) the Somali Cluster(cross-border areas of Ethiopia, Kenya and Soma-lia); and (c) the Afar/Issa Cluster (cross-border re-gion of Djibouti and Eritrea).122

CEWARN has actors at the community, nationaland regional level; field monitors collect data atthe local/community level and are able to informand work with local committees to identify con-flict indicators and resolutions. Data collected isfed into a data monitoring system at the nationallevel based on 52 indicators.123 The data systemproduces alerts as they occur, in addition to pro-ducing monthly updates and quarterly cluster re-ports. More recently, CEWARN has begun to makeuse of crowdsourcing SMS alerts to complementdata collection.124 Conflict Early Warning andEarly Response Units (CEWERUs) at the nationallevel receive the reports and warnings, and areable to engage in response or mobilize communityauthorities to engage in conflict resolution. Data atthe national level is also analysed by National Re-search Institutes that report to relevant agencieswithin the IGAD Secretariat and Council of Minis-ters.125 This chain of information sharing allowscoordinated interventions at the micro-level, thenational level, and in some case cross-nationally.

CEWARN therefore relies heavily on national andcommunity structures, and each CEWERU shouldconsist of a steering committee (made up of civilsociety, members of parliament and security per-sonnel), a focal point, and the local committees ref-erenced above. In 2006, the decision was taken toexpand the mechanism to cover all IGAD memberstates,126 and IGAD’s strategy for CEWARN for2012–19 responded to further demands to signifi-cantly expand its thematic and geographical focusin order to address a broader spectrum of driversof conflict.127 For some states, this remains the goalrather than the reality: in 2015, CEWARN was onlyfully operational Kenya, Ethiopia and Uganda, and

was beginning to be more active in South Sudanand Somalia.128

Cross-border civil society networks have been ef-fective in supporting the work of CEWERUs, suchas in Karamoja. Operational challenges, particu-larly lack of payment for field monitors, who playa crucial role in data collection and have faced fi-nancial constraints in carrying out their roles, hasled to an increased reliance on CSOs for data col-lection.129 Generally speaking, human resourceand funding constraints remain a significant chal-lenge to the functionality of CEWARN across theregion.

Role of civil societyThe preceding sections outlining some of the re-gional systems operating across Africa demon-strate the crucial role that CSOs play in regionalearly warning structures. In fact, without civil so-ciety, it is questionable whether CEWARN andECOWARN, and consequently CEWS, would beable to operate. The reliance placed on civil societynetworks and organizations to provide data, en-gage in data collection, play roles in peace commit-tees and national steering committees, and engagein conflict response at the local and national level,demonstrate the extent to which civil society con-tributions are needed within these systems. With regard to CEWARN, it has been highlightedthat civil society has made significant contribu-tions in the following areas:130

• Capacity building – providing training tosecurity forces, national and localgovernment as well as community leaders.

• Facilitating community dialogue.• Accountability – monitoring conflict

dynamics and peacebuilding efforts enablescivil society to hold state and non-state actorsto account.

• Providing an entry point to communities –CSOs are often closer to local communitiesthan state actors and have alreadyestablished relationships. In fact,communities may well be suspicious of stateor security force attempts to gather earlywarning data and the impact it may have onlocal communities.131

• Expertise – proximity to the local populationmeans that civil society is well placed toassist government stakeholders in designing


22

appropriate responses and identifying theright people in the community to work with.‘For example, this was the case in northernUganda, in areas affected by the conflict withthe Lord’s Resistance Army. CSO involvementin response can ensure response isembedded in communities.’132

The reliance of CEWARN and ECOWAS on nation-alized structures demonstrates that local and na-tional early warning efforts are needed tocomplement regional systems. Where gaps exist incovering conflict-affected areas, leveraging localcivil society actors who have existing expertiseand networks is preferable to deploying field mon-itors. Integrating local, national and regional ef-forts at early warning and conflict prevention canhelp to build a more comprehensive overview of

conflict-affected areas, as well as reduce dupli-cated efforts.

The diversity within many states in Africa meansthat local knowledge is necessary in order to buildeffective early warning systems. The expertise thatcivil society has of local actors, customs and lan-guages means that civil society should also play acrucial role in developing indicators for localizedearly warning.133

Importantly, unlike regional systems, such asECOWARN or CEWARN, local or civil society-ledinitiatives create greater possibilities to ensurethat at-risk communities are actually the receiversof early warning information, and are empoweredto resolve conflict within their communities whenappropriate.


The armed conflict inSouth Sudan In December 2013, two turbulent years after its indepen-dence, the world’s newest country descended into civil war.After violence erupted between troops loyal to PresidentSalva Kiir Mayardit and rival Riek Machar in the capital Jubain December 2013, South Sudan’s army quickly split intorival factions and fighting spread rapidly across the country.

The international community started working towards re-solving the conflict soon after it started. By January 2014,the IGAD, an eight-country trade bloc in East Africa, hadbrokered the first ceasefire, the Cessation of HostilitiesAgreement (CoHA), between parties to the conflict. This wasfollowed by the first peace agreement, the August 2015Agreement on the Resolution of the Conflict in South Sudan(ARCSS). However, the agreement was derailed as fightingerupted between Kiir’s and Machar’s personal guards inJuly 2016 and the country returned to widespread violence.

By December 2017, the international community had bro-kered a second ceasefire, the Agreement on the Cessation ofHostilities (ACoH), which paved the way for the secondpeace agreement, the September 2018 Revitalised Agree-ment on the Resolution of the Conflict in South Sudan (R-ARCSS). Although the formation of a Revitalised TransitionalGovernment of National Unity in February 2020 brought theconflict to a formal end, the peace process is generally rec-ognized to have stalled and the country continues to be af-fected by widespread violence.

While the second ceasefire and peace agreement havelargely held between the parties and prevented a large-scale return to fighting, serious and large-scale outbreaksof violence have continued across the country. Rival fac-tions, some signatory to the second ceasefire agreement,some to the ceasefire and subsequent peace agreement,and some to neither, have resorted to force to defend or im-prove their positions on the ground to give themselves in-creased political leverage. Further, low-level violence isendemic across the country.

Piloting civilian-led earlywarning in South Sudan

The Community Empowerment for Progress Organisation and the Ceasefire

Centre for Civilian Rights have, over the last two years, built and operated a pilot

fourth-generation early warning system for the identity-based violence that has

characterised the non-international armed conflict in South Sudan. It is both

civilian-led and civilian-focused, and associated advocacy has successfully

raised issues of civilian protection and civilian rights with key international actors.

2

24 First to Know: Civilian-led early warning in armed conflict

The conflict has been characterized as ethnic, andindeed initially had an ethnic character as long-time political rivals Kiir and Machar134 looked totheir own ethnic groups for support, but this ethnicorientation increasingly broke down at the locallevel as the war progressed, as local groups mobi-lized and aligned themselves with the warring fac-tions to settle scores or advance their interests.Though the war on the national level has beenfought between a Dinka and a Nuer leader, andthis ethnic divide has retained its salience through-out the conflict, on the local level, the war has pit-ted Dinka against Dinka and Nuer against Nuer.135

Attacks on the civilian population have been amodus operandi for all sides in the South Sudanconflict almost since the war broke out, as a wayto secure advantage on the ground, signalling se-riousness to others while jockeying for positionand access to resources and avoiding the risk ofserious casualties by avoiding engagements withother armed actors. A long-serving member of theinternational ceasefire monitoring mechanismstaff noted to Ceasefire that the conflict had beencharacterized less by clashes between the partiesthan by parallel attacks on undefended commu-nities perceived as belonging to, or supporting,other parties. Such attacks have been character-ized by pillage, destruction of property, killings,sexual violence and mass forced displacement,and starvation has been employed by both sidesas a weapon of war.136

The scale of violence has been horrific. A Septem-ber 2018 study by the London School of Hygieneand Tropical Medicine concluded that at least383,000 had died as a result of the civil war, abouthalf of them directly through violence.137

The violence has also caused mass forced dis-placement on a staggering scale. Around 4 mil-lion, one third of South Sudan’s population, havefled their homes, with over 2.2 million going out-side South Sudan.138 More than 1.35 million wereinternally displaced as of the end of 2019. Ongoingviolence displaced a further 232,000 people in thefirst half of 2020 alone.139

There are no accurate figures for the level of sex-ual and gender-based violence (SGBV) in the con-flict. However, the evidence that is availableindicates the number of women subjected to rape

and other forms of sexual violence, includingwomen and children taken captive as ‘wives’ bycommanders or raped and beaten by multiplefighters, is extremely high. In an outbreak of or-ganized violence in the Central Equatoriaprovince between September 2018 and April2019, which caused 76,000 people to flee theirhomes, the number of women and girls subject toSGBV was almost as many as the 104 killed in theviolence.140 Levels of SGBV have remained highdespite the ceasefire.

Existing monitoringsystems and civilianprotectionUNMISS has been present and operationalthroughout the war. Established by a vote of theUN Security Council on 8 July 2011, the day beforeSouth Sudan achieved its independence, its statedpurpose was ‘to consolidate peace and security,and to help establish the conditions for develop-ment in the Republic of South Sudan’.141 Initially7,000 soldiers and 900 police, it was reinforced inMay 2014 and its mandate reprioritized towardscivilian protection. By late 2020, it had grown to19,056, including 14,038 troops and 1,699 police.142

Of particular significance for civilian protectionwere the creation of Protection of Civilians Camps(POCs), established in or near the bases of UNpeacekeeping forces. Around 200,000 civilianssought protection at seven UN sites throughout thewar, with the majority still in these camps as of2020 even as they are being returned to the controlof the South Sudanese government.143 Althoughthis is only a fraction of the population displacedby the conflict, the protection of vulnerable civiliancommunities in the POC camps has been a majorUNMISS operation. More broadly, UNMISS has oneof the largest human rights offices of any UN mis-sion worldwide, and its work includes humanrights monitoring across the country.

However, UNMISS is not and has never been re-sponsible for monitoring compliance with cease-fire agreements. The Monitoring and VerificationMechanism (MVM) was established by the firstceasefire agreement to monitor the ceasefire,

25

under the aegis of the eight-country East Africantrade bloc the IGAD.

This MVM later became the Ceasefire and Transi-tional Security Arrangements Monitoring Mecha-nism (CTSAMM) after the August 2015 Agreementon the Resolution of the Conflict in South Sudan(ARCSS), and subsequently the Ceasefire andTransitional Security Arrangements Monitoringand Verification Mechanism (CTSAMVM) after theSeptember 2018 Revitalised Agreement on theResolution of the Conflict in South Sudan (R-ARCSS). This mechanism, in its various iterations,operates separately from UNMISS and has in prac-tice been the only early warning network that hasoperated throughout most of the South Sudanesecivil war.

CTSAMVM and its previous incarnations havebeen chaired by a representative of IGAD. TheMVM, the first iteration of the monitoring mecha-nism, reported to the IGAD special envoy to SouthSudan; its two subsequent iterations have re-ported to the Joint Monitoring and EvaluationCommission (JMEC) and its successor the Recon-stituted Joint Monitoring and Evaluation Commis-sion (RJMEC).144 CTSAMVM’s board currentlyincludes representatives from the AU, China, theUSA, the UK, Norway and UNMISS, as well as somecivil society representatives.145 CTSAMVM hasover 150 staff and runs nine monitoring and ver-ification missions, which can be deployed in crit-ical regions at short notice.146

Though the main task of CTSAMVM and its prede-cessors has been monitoring the ceasefire be-tween the warring parties, its mandate issignificantly broader. It is also responsible formonitoring prohibitions on the recruitment ofchild soldiers and SGBV; the recruitment of newforces, and the agreement of the parties to facili-tate humanitarian access, freedom of movementand access for CTSAMVM personnel; and respectinternational law and the human rights of civil-ians and ensuring the safety and dignity of indi-viduals and communities.147

There has been a large degree of continuitythrough the various incarnations of the ceasefiremonitoring mechanism, however it has evolvedsignificantly through its various incarnations. Itspersonnel generally became better trained and

more capable, and the operation itself becamemore focused on its mandate. However, none ofthe various iterations of the monitoring mecha-nisms were fully integrated into a broader strat-egy to end the conflict, and no attempt was madeto engage or win the support of the public for themonitoring mission.148

From the outset, it was not clear to personnelwhether the monitoring mechanism should focuson reporting violations to international decisionmakers quickly, or on detailed reports of incidentsthat would be of use to international legal mech-anisms. The mechanism appears to have fallenshort on both counts.149

Further, the monitoring mechanism has been de-nied access to certain areas by the parties to thepeace agreement, including the South SudaneseArmy.150 The work of the monitoring mechanismhas also been subject to interference from the me-diation team as they attempted to bring the partiesto the conflict together. Given these shortcomings,and the unwillingness of the international com-munity to hold the warring parties to account fortheir violations, the monitoring mechanism ar-guably failed to establish red lines for any of thewarring parties and failed to deter either viola-tions of the ceasefire or those committed againstcivilians.151

The Ceasefire/CEPOearly warning project Unlike CTSAMVM, the EWS piloted by CEPO andthe Ceasefire Centre for Civilian Rights is civilian-led and civilian-focused. This has meant that fromthe outset it has not been subject to the same in-stitutional political pressures as CTSAMVM, and isconcerned with violence targeting and/or wit-nessed by civilians rather than fighting betweenarmed factions. In the context of a civil war inwhich factions have routinely targeted civilians,152

this distinction is perhaps less clear in SouthSudan than in some other conflicts. However low-level violence, whether criminal or communal innature, is a very real concern for civilians inSouth Sudan and consequently a key focus of re-porting for the civilian-led and civilian-focusedCEPO/Ceasefire EWS, though not for CTSAMVM.


26

Civil society in South Sudan is fragmented, under-resourced and largely politically marginalized.Difficult economic conditions mean that many ac-tivists need to rely on other sources of income,and that organizations have trouble securing theeconomic resources they need to develop and re-tain a cadre of professional staff. Relatively feworganizations operate beyond their immediate lo-cality. CEPO works on a range of issues – capacitybuilding for CSOs, security sector reform, justiceand accountability, conflict mitigation, women’sempowerment, sustainable livelihoods, good gov-ernance, human rights and the rule of law – work-ing towards democratic transformation.153 Thisversatility has probably helped it access fundingfrom international donors, and allowed it to buildup its operational capacity beyond that of mostother South Sudanese CSOs. Its relative closenessto the international community has provided itwith a degree of protection.

CEPO has been operational throughout the civilwar, and has run a conflict map on their websitesince 1 January 2015, collating media reportingabout casualties in the conflict.154 It is well con-nected with the international and diplomaticcommunity in South Sudan, and within South Su-danese civil society. Though it is based in Juba, ithas coordinators in seven cities across the coun-try – Bor, Yei, Torit, Rumbek, Terekeka, Wau, Yam-bio – that work on a range of issues, includinghumanitarian concerns and human rights, and iscurrently in the process of expanding its networkof coordinators across more of the country. Inshort, it was well placed to be the locus of the civil-ian-led EWS.

CEPO has run the EWS from its main office in Jubawith Ceasefire’s active support. The network wasinitially built on CEPO’s existing civil society con-tacts, but quickly expanded beyond that to in-clude journalists, transport companies and otherconcerned civilians. CEPO’s coordinators were en-gaged in the project early on, but as building trustand securing the engagement of various groupshas proven to be time-intensive, it has proven tobe more practical to trial engagement with newgroups from the Juba office first, and use the in-sights and experience gained from these initial tri-als to guide CEPO’s regional coordinators in theirengagement with local groups who could be re-cruited as monitors.

CEPO and Ceasefire elected to avoid the use ofcomplicated metrics and use civilians’ own per-ceptions of threat and violence: the threshold isan incident a civilian deems worthy of reporting.Incidents are examined for credibility, andwhether they might signify any underlying trendsor developments. New developments, patterns orsignificant events detected by the network are im-mediately flagged with CEPO senior leadershipfor targeted advocacy on mandate holders andother duty bearers who have the capacity to un-dertake preventive action. This advocacy hasemerged as a key element of the project and con-tributed to its success in several ways.

Though civilian protection is a key element of UN-MISS’s mandate, civilians effectively have had lit-tle input into the peace process. As expected,advocacy has increased the salience of civilian in-terests and civilian protection among key actorsin the peace process. The partnerships withIGAD’s early warning network Conflict EarlyWarning and Response Mechanism (CEWARN)and engagement with AU early warning mecha-nisms likely gave the system credibility in the eyesof key international actors and facilitated engage-ment with them. This engagement also gave CEPOaccess to secure servers for the storage of infor-mation outside of South Sudan.

CEPO furthered this engagement with key interna-tional actors by using the information generatedfrom the reporting network to facilitate routine ac-cess to these key actors, including senior officialsin UNMISS, by passing on critical information di-rectly via personal rather than formal channels,which probably would have been less responsive.Along with assistance provided in developing andtargeting advocacy messaging, the result has beenthe increased credibility and relevance of civil so-ciety voices in key international decision-makingforums, and culminated with the head of CEPO,Edmund Yakani, being invited to address the UNSecurity Council in its 23 June 2020 session onSouth Sudan. This increased visibility and impacthas resulted in routine access to key internationaland national decision makers, including ambas-sadors and embassy staff, the head of UNMISS andhis staff, and the vice president.

The close contact CEPO has had with key interna-tional actors has also provided a degree of political


27

cover and thereby protection for the organizationfrom potential harassment or persecution fromthe government. Many journalists and activists inSouth Sudan have been harassed and targeted fortheir peaceful opposition or for what has been per-ceived as criticism of the government.

The project has also highlighted the potential of na-tional and international media as a tool for influ-encing decision makers. Journalists have beenengaged by the project as monitors, but contactswith media have also proven useful to drive advo-cacy. On 3 January 2021, CEPO released a statementto the media revealing that its EWS had recordedover 2,450 civilian deaths and calling on the gov-ernment to take action to halt this.155 The statementcontributed to governmental action. On 5 Januarythe government convened a meeting of the Na-tional Security Committee and directed the securitysector to ‘immediately put a halt to road ambushes,cattle raids, and inter-communal clashes in thecountry’.156 The issue was also covered in Voice ofAmerica’s South Sudan radio edition. The govern-ment’s prompt response to this media coverage islikely to be a reflection of the success in getting theissue covered in international media – but also ofthe contacts and credibility CEPO has built with theinternational community and its consequent abilityto influence their concerns.

Further, this engagement with key actors has alsoproven to be critical in the development of theearly warning network. These contacts have giventhe project credibility and – most importantly –relevance in the eyes of civilians. From the feed-back we have received and the patterns of en-gagement we have observed, we have concludedthat civilians are generally very pragmatic in theirapproach to early warning systems. They will en-gage if they think it is worthwhile – that is, if it isseen as a thing that could have a real impact ontheir lives. A network that generates reports of vi-olations committed against civilians that are thenfired into the electronic ether and produce no tan-gible results for the civilians themselves is simplynot worth their time.

Demonstrating the relevance of this EWS to civil-ians has been a challenge, as different individualsrespond differently, based on their own knowl-edge and engagement with wider issues. For in-stance, civilian activists who, by definition, are

trying to achieve change more readily relate tothe aims of the project and have generally beenwilling to participate. They understand what theproject is trying to achieve, and see its potentialrelevance to their lives and their activism. In con-trast, business-people see the potential of the pro-ject, but want to assess its likely benefits – andrisks – for their work before they commit. As a re-sult it has taken significantly more time and effortto secure their active participation.

The potential risks associated with participationin the EWS have been another significant consid-eration for many. CEPO has had to engage withthese concerns and demonstrate that they can beameliorated in order to secure the participationof many civilian monitors. In the context of SouthSudan, security is potentially problematic. Allarmed factions to the conflict have all been ac-cused of committing serious violations, and sev-eral activists and opposition figures have beentargeted for their peaceful activities. Civilian par-ticipants in the EWS have had legitimate concernsabout sharing what might be, for factions in-volved in the conflict, sensitive information.

CEPO has found it best to raise the issue of secu-rity at the beginning of any prospective relation-ship with civilian monitors and to revisit thisissue whenever necessary as the relationship de-velops. It has proven to be neither been possiblenor desirable to detach security from the overallrelationship between CEPO and monitors andtreat it as a separate issue. CEPO discussed secu-rity and communications with each potentialmonitor in its early warning network from theoutset, and developed a communications protocoltailored to every participant. Some did not havereliable access to the internet, or even mobilecommunications; others did not trust phones andinitially preferred to communicate face to face.

In South Sudan, the landline network is virtuallynon-existent across much of the country, the mo-bile network is unreliable in the limited areas itdoes serve, which are generally centred on themajor cities. Most people even in these urbanareas do not own mobile phones, and data andminutes are expensive. The mobile communica-tions app WhatsApp is commonly used and, de-spite potential security flaws which the SouthSudanese government may have the potential to


28

exploit, it is popular among activists because of itsconvenience and its suitability for messaging andeven voice calls in areas of poor or sporadic mo-bile internet coverage. Alternative and more se-cure messaging apps like Signal have proven to beless reliable for voice calls, which are preferredto messages as a means of transmitting sensitiveinformation. Email is used in professional con-texts, but is less common for routine and personalcommunication.

Most respondents were likely to make only occa-sional reports, while a few (see below) were inroutine contact. The communications protocolsreflected this. A key issue that has emerged ismaintaining contact with the network; withoutroutine contact, participants may forget or losemotivation and fail to report incidents. This hasnot been as significant a problem as was origi-nally anticipated, as it has been dealt withthrough routine communications protocols byagreeing to touch base occasionally, and routinecontact has been a key element in building trustwith participants.

Where serious incidents have been reported thatdirectly affected participants, follow up has beencritical in maintaining trust. In most cases, thishas consisted of informing UNMISS and local au-thorities and following up with them to ensurethey take action, and communicating this to par-ticipants. Even when UNMISS and local authori-ties have not been willing or able to respondeffectively, CEPO’s efforts and the efforts it hasmade in following up have generally been appre-ciated by participants, though these occasions in-evitably undermined confidence in theeffectiveness of the EWS.

CEPO has successfully engaged individuals froma range of social, activist and business groups asmonitors, including civil society and women’s ac-tivists, religious groups, CSOs and business-peopleworking in the transport sector. Because of lim-ited resources, and the time it has taken to de-velop networks of contacts and refinemethodology, the project was initially focused pri-marily on the Juba region and is currently seekingto expand across more of the country. The follow-ing two subsections present two case studies onCEPO’s engagement with different groups to bringthem into the EWS.

Case study 1: Transport companies An interesting case study of how this relationship-building can proceed is the CEPO/Ceasefire sys-tem’s engagement with the public transport sector.It was realized early on that transport companiesregularly cover large areas of the country, movingboth people and goods, and have a vested interestin the safety and security of their routes – and es-pecially of their staff, vehicles and passengers.

South Sudan’s public transport sector has beensignificantly affected by the civil war. Before thewar, fewer, larger companies ran a network ofnational and local routes through urban trans-port hubs. Because of the disruption and violencethe civil war brought, these have now frag-mented into many different companies operatingfrom urban hubs running single routes. Theseroutes generally run between urban centres andmay stretch as far as 150–200 km in different di-rections.

CEPO first reached out to transport companies op-erating from Juba. Though the managers of thesecompanies responded well to initial approaches,it took a long time for them to fully trust the ini-tiative, understand what they needed to do andwhat they could expect from it, and work out a vi-able modus operandi. We estimate this took threeto four months in total from initial contact to con-structive engagement.

Though the initial contact with the transport com-pany managers was done collectively, we foundthat to actually secure their cooperation it wasbetter to deal with each manager individually –they were more willing to discuss details andcommit to participation in confidence. Even inprivate, it took regular visits over weeks for man-agers to take the proposal seriously and begin se-riously to consider what it would mean for them.As the managers got to know the CEPO staff mem-ber tasked with engaging them, their trust in-creased. Though most managers were willing toparticipate, some were more enthusiastic thanothers and a minority flat out refused to cooper-ate without a significant financial incentive.

It also took them time to understand that UNMISSforces would not always respond in the ways theywanted or expected – that we could raise issueswith UNMISS and follow them up, but that a deci-


29

sive response on the ground was not always forth-coming.

The communications protocols CEPO used to con-tact the managers of the transport companies alsodeveloped considerably. Initially, many of themanagers were not happy communicating infor-mation on the phone, and the responsible CEPOstaff member agreed to visit the transport man-agers once a week, or after a phone call in theevent of a serious incident. After a few months,many felt secure communicating over the tele-phone, but CEPO has had to pay them smallstipends to cover the cost of phone calls.

CEPO is now working to engage transport compa-nies in other cities, where its expanding networkof coordinators is located. However, it is likely totake some time before these transport hubs areable to contribute effectively to the early warningsystem, as the same slow process of winning trustand establishing the relevance and value of theEWS will need to happen with each potential par-ticipant, and CEPO coordinators also need to bebriefed and supported as intermediaries in the de-velopment of these relationships.

Case study 2: Women’s groupsCEPO has also had success in engaging with net-works of women’s groups and activists, althoughits experiences engaging these groups alsodemonstrates the difficulties that can be encoun-tered in building civilian-led monitoring net-works.

CEPO reached out to women’s activists early in theproject, and through them to the networks of ac-tivists and small organizations they were in con-tact with, which were mostly concentrated in Jubaor neighbouring regions. Many of these activistsand organizations had not had a great deal of con-tact with CEPO previously, but were happy to passon information and raise concerns through atrusted intermediary who knew CEPO better. Spe-cific cases were followed up by CEPO staff to ob-tain more information.

Unfortunately, communications with the networkof women’s activists became more difficult andthe flow of information decreased when a keycontact and supporter of the project left her posi-tion. CEPO is currently trying to build relation-

ships with new contacts who will support the EWSmore actively.

The particular focus of these activists and groupswas sexual and gender-based violence (SGBV). Thisinformation was passed on to CTSAMVM and UN-MISS, and although the latter demonstrated an es-pecial interest in SGBV-related information, itproved difficult to get feedback on what either ac-tually did with this information. On one occasion,patrols were established in an area of informalhousing near Juba in which women had been sex-ually assaulted. A police protection unit was alsocontacted about persistently high levels of sexualassault on women, and though they did evince con-cern their limited budget meant they were not ableto extend their operations into the relevant area.

Lessons for civilian-ledmonitoringCEPO and Ceasefire’s experience has producedvaluable insights into what considerations shouldbe taken into account when designing and run-ning civilian-led early warning systems. These arethe importance of understanding and adaptingthe project to its context; considerations of re-sources, organization and structure; and the im-portance of building trust and responsiveness tocivilian concerns.

The importance of context An understanding of context is of critical impor-tance in the design and development of any EWS.For the sake of clarity here, considerations of con-text will be divided into three separate sections:the modalities of conflict; society and civil society;and the broader political context.

Modalities of the conflictThe modalities of the conflict are relevant becausethey directly concern who and exactly what theEWS will monitor. The objectives of the forces en-gaged in the fighting, and the ways the factions mo-bilize forces and act to achieve those objectives arecritical in determining how an EWS should be or-ganized and work. Further, the drivers and natureof the violence cannot be assumed to be the samethroughout the conflict zone, and allowances forthis variation must be made in the project design.


30

An international expert-led EWS with a limitednumber of observation teams is well-suited to aconflict in which the scope of the conflict is geo-graphically limited and military forces square offprimarily against each other across an establishedfront line from which civilian populations havehad the opportunity to flee.

However, this situation is rare in modern con-flicts, which often take place within the bordersof a state in response to a serious rupture or col-lapse of a political system, and which are oftenmore fluid and may involve the systematic target-ing of civilian populations.

In South Sudan, the conflict has been fought by anassortment of formal and informal forces, withlocal fighters largely mobilized by localgrievances, fears, ambitions and the promise ofopportunity to loot becoming more important asthe war progressed. The diffuse and expansiveroutine coverage over large areas that civilian-ledearly warning networks have the capability toprovide is ideally suited to a conflict such as theSouth Sudanese civil war, which is characterizedby episodic outbreaks of violence, with highly mo-bile armed groups sweeping across the countryand targeting civilians perceived to belong to orsupport opposing groups as a main objective ofmilitary action.

However, the conflict in South Sudan does posechallenges for an EWS. The informal and localizednature of the conflict, coupled with the situationof the at least partial collapse of state authorityand institutions across much of the country hasmeant that it can be a challenge for advocacy toprotect civilians to achieve real change. The cen-tral government authorities may not be directlyinvolved in localized fighting, and the absence ofofficials on the ground may mean the governmenthas limited ability to take effective action, andthough the UN has a significant presence in thecountry its resources are in practice limited withregard to the scale of the crisis it is facing.

Society and civil society Another important element of a civilian-led EWSis the nature of society and civil society in thecountry in question, which determines who canbe engaged to monitor violations and their moti-vations for doing so.

To win the active support of monitors, the EWS hasto win their trust and demonstrate its relevance.Both take time and effort, but of the two demon-strating the relevance of the EWS may be the moredifficult, as it requires proof of the EWS’s effective-ness, and proof that it can actually make a differ-ence in people’s lives. This is why active andongoing engagement with monitors is needed toensure that they understand what they are doingand why, and see the difference the EWS can maketo their lives and the lives of other civilians.

Monitors should be independent or, at the veryleast, have a degree of autonomy from the war-ring factions. They should not identify with thewarring factions and their interests should be dis-tinct – security at the local level, not the victory ofa faction in the war. In conflicts that developalong ethnic or religious lines this distinction maybe less clear, but in most of these situations vio-lence between ethnic groups drives populationsto seek protection with their own ethnic group orco-religionists out of necessity, not choice, and be-yond their need for protection they may not par-ticularly identify with the interests of any givenfaction. In this case, monitors should be selectedwho do not have any close family or other associ-ation with the groups engaged in the conflict, andtheir reliability and impartiality should be as-sessed over time against other independentsources of information.

The possibility for the manipulation of the EWS tothe benefit of one of the parties to the conflict canbe further addressed by developing multiple linesof reporting, with multiple monitors reporting fromthe same area. While many incidents will be local-ized and may only be reported by one monitor,major incidents and serious escalations in fightingare likely to affect a large area and be reported bymultiple monitors. There is also the possibility of in-dependent confirmation of incidents, through othermonitors or the coordinators of the monitoring net-work, when they have the possibility of reachingthe location of the incident or accessing other wit-nesses. However, this requires a considerable effortthat is likely to only be justifiable when it is partic-ularly important to obtain more information orconfirmation about the incident.

In some cases it may not be possible to verify a re-port that does not come from a trusted source, es-


31

pecially, for instance, if the source of the informa-tion is a witness or a victim of violence. The infor-mation and the source must be assessed quicklyfor credibility – whether the account of the viola-tion is consistent both in itself and with the EWSteam’s knowledge of the wider conflict, andwhether the report could be motivated by politics,revenge or other malicious intentions.

However, given the time-sensitive nature of the in-formation produced by the EWS, the passing on ofinformation should not be delayed unless there isgood reason to be suspicious of it. Where furtherinvestigation is necessary, this can be released lateras an update to the initial report. The EWS teamshould employ a dynamic assessment process,using lessons learned from continuous assessmentof its performance to develop better judgement asto where the balance of probability lies in any re-port. The inclination of parties to the conflict to in-terfere with reporting should of course be assessedand factored into this assessment. We have beenconstantly aware of the possibility of malicious in-terference from various actors in the operation ofthe CEPO/Ceasefire EWS, but we have yet to en-counter a single verifiably false report.

Civil society, journalists and business can alsoplay an important role in early warning. Civil so-ciety groups may have the advantage of beingwell connected – with other civil society groupsand the wider society – and may have a trackrecord of activism which can be reviewed to as-sess their impartiality. Journalists may already beinvolved in covering the conflict and may alreadyhave their own network of national and local con-tacts, and any biases and partialities will probablybe easily discovered through their previous re-portage. Journalists covering the conflict can becontacted and engaged as potentially usefulsources of information, and may indeed reach outto the EWS of their own accord. However whenengaging journalists it may be necessary to reachan agreement beforehand on what, if anything,they can publicly report about the network.

Business-people may also be able to make a valu-able contribution, especially if their business in-volves moving goods or people across the country.Businesses may of course be compromised by ac-tual or potential ties to warring factions – and manybusiness-people have indeed profited from war in

one way or another – but conversely many busi-nesses have a real stake in peace and security. Busi-ness-people should be assessed for possibleconflicting interests and compromising ties, andtheir reports may be checked against other sources.

Though the conflict in South Sudan has been seenas being fought between ethnic groups, therehave also been significant divisions withingroups. The most prominent national leaders be-longing to each group do not and have never spo-ken for or even represented the interests of allmembers of that ethnic group,157 and the politicaleconomies that sustain the warring factions havelikely never benefited the majority of the ethnicpopulations they claim to represent. This meansthat, for instance, the assumption that all or evenmost Dinka support Kiir in any meaningful fash-ion is certainly false. This means that it is proba-bly safe to assume that those with no directinvolvement with or relationship to those in-volved with a warring faction most likely have avested interest in peace and security in their ownlocal community. Even those who have benefitedindirectly from the conflict, such as populationsthat have gained from the ethnic cleansingthrough the forced displacement of rival ethnicgroups, can be assumed to have an interest inpeace and security on the local level.

Further, though South Sudanese civil society isunder-resourced and underdeveloped in thesense that there are relatively few dedicated andprofessional CSOs, there are many civil activistsand many community-based groupings or associ-ations that perform many of the functions of civilsociety at the local level. These groups haveproven to be interested and willing to engage withthe EWS, and keen to see if it can leverage realchange in their lives at the local level.

Advocacy and the broader political situation Advocacy is critical for the success of the project.The EWS is collecting information to make a dif-ference to people’s lives, and it needs a viable ad-vocacy strategy to be able to influence keydecision makers to do so.

This means the EWS must engage effectively withthe broader political context in which it operates.In most though not necessarily all cases, this willmean both the domestic political context but also


32

engaged international actors, which will includeforeign states, the UN and possibly other interna-tional institutions. The advocacy strategy must bebased on sharing the information produced by theEWS, and the insights into grassroots violence andthe concerns of the civilian population that can bederived from it, which can be used to leverage ac-cess to key decision makers. In some instances,simply sharing information may be enough, butin most cases the information needs to be craftedinto an argument for action that is concrete anddoable. Specific recommendations based on infor-mation about specific incidents is likely to be mosteffective. General calls for action will be more ef-fective when backed by statistics and specifics.

Political developments in the country and in thecountry’s relationship with the international com-munity may directly or indirectly impact on thework of the EWS by changing the parameters orcontext within which it operates. It is importantto recognize that, while early warning systemsmay be apolitical in their design and intent theyoperate in inherently political contexts, and par-ties to the conflict or violence they monitor maysee them as politicized and attempt to underminethem, or instrumentalize them in their wider po-litical strategies. This means it is essential to fol-low all developments in a country and analysethem for potential impact on the project, and foropportunities they might offer to increase the ef-fectiveness and impact of the project.

The operation and possibilities for action of an EWSwill also depend in part on the political and institu-tional situation that mandated the establishment ofthe system. Those managing the system must beaware of this and factor it into their calculations ofthe political context in which it operates. Those op-erating an EWS must be prepared to negotiate redlines and operational limits set by key national andinternational actors, and understand when andhow to manoeuvre around them to keep faith withthe people the system is meant to protect and onwhose cooperation it ultimately depends. This willbe especially true if the EWS was imposed on a na-tional government by the international community,or approved grudgingly by a national government.

It must not be assumed that all national and inter-national actors will act in good faith, and thepeace or resolution to the conflict they claim to

support may reflect their interests rather thanthose of civilians in the conflict zone. Parties tothe conflict may also attempt to instrumentalizeinternational efforts at mediation, or to support apeace process to pursue its war aims through theinternationally sponsored peace process. Foreignstates involved in the peace process may also havetheir own interests in securing particular out-comes to the conflict. However, even when indi-vidual international actors may be biased, theengaged international community as a whole maybe able to exert a positive influence and be a use-ful interlocutor, providing useful support for acivil society-led early warning system.

Further, monitors or a CSO running the networkcould be targeted by parties to the conflict forwhat they perceive as negative or critical report-ing on their activities. A strong relationship be-tween the EWS and impartial national orinternational actors willing to intervene andbring real pressure to bear on those who threatenit will offer protection against this possibility.

The media may also prove to be useful for educat-ing and engaging the general public, possibly evenfacilitating engagement with new monitors, andto influence and put pressure on key decisionmakers. Journalists and international correspon-dents may be useful contacts for an early warningnetwork, but they can also provide potentiallyuseful coverage of violence at the local or nationallevel to influence key decision makers. When in-fluencing foreign diplomats, reaching out tomedia from their own country is likely to be par-ticularly effective.

Luckily opportunities for impactful advocacy areplenty in South Sudan, as the effort to resolve theconflict has become thoroughly internationalized.The international community has been engagedin negotiating an end to the conflict in SouthSudan since the first year of the war, and cur-rently the UN, the AU, IGAD and an assortment ofstates from Africa and the wider world are in-vested to some extent in the peace process.

Further, and critically, after years of brutal anddestabilizing conflict, the predominant interest ofregional states in the South Sudanese civil war isthat it ends and brings with it some form of sta-bility. The regional rivalries that initially raised


33

fears the conflict could ignite a wider war haveabated and interest is now focused on ending thewar. This is a sharp contrast to, for instance, thecivil conflicts in Syria and Libya, which have alsobecome proxy wars for regional and internationalpowers. In South Sudan the intractable nature ofthe conflict itself has prevented its resolution, notregional and geopolitical rivalries.

However, the civil war has its own particularchallenges. The South Sudanese state has partiallycollapsed and state authority and institutions arelargely absent across much of its territory. Localarmed groups, who may be aligned with majorfactions but are not in any meaningful way undertheir direct control, are present in many areas.

Furthermore, there is a strong argument that, inits engagement with the peace process, the gov-ernment of South Sudan has not acted in goodfaith. There is evidence it has been both directlyand indirectly involved in violence against civil-ians during the civil war; that it has acted to im-prove its position at the expense of civilians,preventing access of monitoring teams to keyareas of the country and commissioning or en-couraging violence that has forced hundreds ofthousands to flee their homes; and that its engage-ment in the peace process has been less than en-thusiastic and whole-hearted.

The engagement of the international communityhas been extremely useful as a counterweight tothe government. CEPO has intentionally culti-vated close ties with the international communityto provide protection against possible retributionfrom the government for its activities, and hasused the early warning system to develop an evencloser relationship with key international actorsincluding UNMISS, the AU, IGAD and certain for-eign embassies.

Because of this situation, the primary recipientsof information from the EWS have been the inter-national institutions – UNMISS, and the AU andIGAD early warning networks. Information hasbeen shared with the South Sudanese authorities,but on a case-by-case rather than a routine basis,for instance when local police were engagedwhen dealing with incidents of violence againstwomen. The credibility and perceived influencethe EWS has on its international contacts has

meant that local authorities have taken issuesraised with them seriously.

Running a civilian-led EWS:practical considerations The key considerations that must be taken into ac-count when running a civilian-led EWS are three-fold: the management of its operations, security,and broader issues of strategy and development.

Operational management The operational elements of an EWS can be orga-nized and run in a variety of ways as long as theymeet certain strategic criteria essential to the suc-cess of the project; primary among these is that itis suited to the environment in which it operates.

The critical operational element of any civilian-led EWS is civilian outreach and engagement. Thisshould be handled differently depending on thesociety and the groups that will be engaged,whether these are business, the media, civil soci-ety and activists. It is to be assumed that there aresignificant differences between how these sectorswork and are structured in different countries,but a civilian-led EWS must go beyond thesebroad differences and deal with society on themost granular level – at the level of individuals.

It must be recognized that reaching out to, engag-ing and keeping individuals engaged involves asignificant and ongoing investment of time. Thereare no shortcuts to this that do not pose a signifi-cant risk of undermining the project.

While engagement with each monitor or potentialmonitor should be conducted within the broadparameters of the project, each monitor needs tobe engaged individually; a bespoke informalworking agreement needs to be arrived at witheach, and this needs to evolve with their situation,the situation in their locality and the situation inthe country. Routine communication with moni-tors, perhaps on a weekly or monthly basis, al-lows for potential changes to be monitored andassessed, as well as providing reassurance formonitors and making an argument for the rele-vance of the EWS to their lives.

Modalities of communication need to be devel-oped to suit the situation of every monitor, and re-viewed and developed as required to suit their


34

evolving situation. In general, monitors in SouthSudan felt more comfortable starting with a moreformal and structured communications protocol,and as their familiarity with and confidence in theproject increased, communications protocolsevolved to become less formal and structured. Formany monitors, the cost of communications maybe significant and stipends to cover communica-tions costs should be paid where this cost is notnegligible. In addition to facilitating and incen-tivizing communication, stipends to cover the costof telecommunications also serve to demonstratethat the EWS is serious and willing to invest in itsmonitors.

Similarly, the nature and quality of the informa-tion gathered by the EWS from civilian monitorswill vary significantly. This will require flexibilityin how it is processed, stored and analysed.

Provision must also be made to follow up on indi-vidual cases when required. The nature of this fol-lowing up will vary, depending on the violationand the overall situation – whether it is judgedmost appropriate to raise the case with local, na-tional or international authorities or some combi-nation thereof. This might involve pressure tobring those responsible for violations to justice, orit might involve preventive action, for instancefor police or peacekeepers to take action to pro-tect civilians from violations or deter potential vi-olators. This is important not just in terms ofsecuring justice for those who have suffered seri-ous violations, but also to maintain the credibilityof the network among those who stand to benefitthe most from it and without whose willing andactive support it cannot function.

A process of dynamic assessment should be em-ployed to analyse information as it comes in. In-formation should be assessed both in itself and asit relates to the broader phenomenon of violencein the country to identify patterns – is it an iso-lated incident, part of a general pattern or per-haps part of a wider incident like an armed groupon the move? This should be reflected in advocacymessaging, and for this reason analysis and advo-cacy should overlap and not be entirely separatefunctions within the EWS team.

Another significant operational concern is re-sources and staffing. Running a civilian-led EWS

requires the investment of a great deal of timeand energy. Building a network takes time, as it isnecessary to reach out and engage, to explain andwin trust, and – critically – to demonstrate the rel-evance of the EWS to the lives of civilians. This isan ongoing process, and some version of it has tobe repeated for every participant. As civilians re-locate to other locales, leave their positions inCSOs, or even leave the country, it will be neces-sary to recruit new monitors and the entire pro-cess will need to be repeated. Expanding themechanism into new localities may be even moretime consuming, as this may require developingnetworks of new contacts from scratch, and tap-ping into existing networks of CSOs and activists,winning their confidence and securing their buy-in to the project takes a lot of time and effort.

SecurityA strong commitment to the security of monitorsis critical for the success of a civilian-led EWS.This includes secure communications protocols,secure information storage and anonymizing in-formation as much as possible before it is re-leased to decision makers. Further, and as notedabove, a close relationship with impartial na-tional or international actors willing to act to de-fend the EWS and its monitors will mitigate thepossible dangers for monitors and staff.

As the risks and the security concerns of eachmonitor vary, so must the security measurestaken in communicating with them, and in usingand anonymizing their information. This shouldbe discussed openly with each monitor at the be-ginning of their relationship with the EWS, andraised with them routinely after that, and in re-sponse to any emerging potential threats to theirsafety.

Communication protocols should be allowed toevolve as relationships with monitors develop,and their situation and the situation in the coun-try changes. As familiarity with the project in-creases, and trust and confidence develop,communications protocols can be revised to makethem less rigid and demanding.

As monitors live and work in zones of active inse-curity or conflict, it can be assumed that they willhave an acute awareness of the risks and the stepsthey may take to mitigate them, and these could


35

of course include reporting these risks to the EWS.Coordinators should regularly discuss monitors’personal security with them and see if they cangive them any advice or perhaps if there is train-ing available for them. Information in conversa-tions with monitors about security should flow inboth directions, from monitors to coordinatorsabout the situation they are experiencing on theground, and from coordinators to monitors aboutpotential or emerging threats they are aware offrom other monitors or at the national level, forinstance if the security of channels of communi-cation is potentially at risk.

Information storage and properly anonymizinginformation is also a very real concern. No matterhow securely encrypted information on a harddrive is, it is vulnerable if people who have accessto it are arrested and tortured or otherwise pres-sured to reveal passwords, and disk drives physi-cally present in the country can also be seized orstolen. For this reason, it is desirable to store in-formation outside of the country where possible,and to have several individuals in the organiza-tion, and even outside of the country, who canshut down access if others are arrested or other-wise compromised.

The anonymizing of data to protect sources is im-portant, though there may be some tension be-tween the desire to protect sources and thenecessity of revealing information. These must behandled on a case-by-case basis and, when it isnecessary to reveal information that may possiblyendanger a source, discussed with monitorswhere possible. Their decision as to how much toreveal should be accepted as final.

Security should be viewed as a dynamic elementof any EWS, and security protocols should be con-stantly reviewed and adjusted to reflect the devel-oping or emerging threats, whenever possible inan active dialogue with monitors.

Strategic developmentBecause of the difficult circumstances in whichearly warning systems operate on the ground,and the challenging and dynamic advocacy envi-ronment in which they function, it is of critical im-portance that mechanisms for feedback andimprovement be incorporated into them fromtheir inception.

Further, a civilian-led EWS must invest in and ac-tively seek to develop the human capital on whichit depends, improving the skills, knowledge andawareness of monitors and EWS staff, and strength-ening the relationships between monitors and staff.Given the varied and potentially far-flung situa-tions in which monitors operate, it is likely that at-tending formal training sessions will be difficult, soit will probably be most effective for EWS handlersto view their routine interactions as possible oppor-tunities to impart knowledge and expand monitors’knowledge – of human rights, the law of war, stan-dards of evidence, national and international ju-risprudence, relevant international institutions, thelatest political developments and/or anything elsethat is relevant. This education should begin withthe initial engagement with monitors, so they startto understand what they are doing from the verybeginning. The investment of this time in the mon-itors will likely increase both their value to the EWSand their commitment to it.

As advocacy is an important element of any EWSproject, an investment in increasing the relevanceand impact of the messaging of in-country civil so-ciety partners, who will most likely be the pro-ject’s interlocutors with key decision makers, willpay dividends. As their credibility and relevanceincreases, their advocacy work related to the EWSwill become more impactful.

Further, and critically, advocacy strategies needto be constantly reviewed and developed as thepolitical environment, policies, personalities andopportunities for advocacy can change rapidly.Advocacy plans, key targets and messaging allneed to be constantly reviewed in light of currentdevelopments. Though Action Research, whichenvisages a cycle of planning, action and assess-ment, is perhaps the most common conceptualiza-tion of a feedback mechanism, the author hasfound the OODA loop (observation, orientation,decision and action)158 more useful. Its stress onorientation in light of recent developments andconscious decision making is well-suited to com-plex, interdependent, rapidly changing situations.

Trust and responsiveness to civilian concerns The most important elements of any civilian-ledEWS are winning and maintaining the trust of thecivilians on which it ultimately depends, and es-


36

tablishing its relevance to their lives and the livesof other civilians.

Civilian monitors have to trust that the EWS, itsstaff and the organizations running it will do whatthey claim to do – represent their interests and theinterests of other civilians honestly and diligently.They also have to believe that these efforts will berelevant to their lives, that they can actuallyachieve or credibly claim to aspire to meaningfulchange.

The problem for the EWS is that trust and credi-bility is hard to win, but easily lost. In practicewhat this means is that mistakes, missteps oromissions on the part of the EWS in dealing withits civilian monitors and the incidents they reportwill undermine trust and the credibility of theEWS more than its successes will bolster it. An ap-parent lack of concern for the personal securityof the monitors will be particularly damaging fora civilian-led EWS.

Given that missteps in any complex EWS opera-tion are all but inevitable at some stage, thismakes routine, open and honest communicationwith monitors and the active management of ex-pectations critically important. When mistakesare made or shortfalls exposed, EWS staff need tobe able to discuss them frankly, and should re-spond with honesty and clarity, and, where nec-essary, with urgency to monitors’ concerns. Thecareful management of relationships with civilianmonitors is essential if disillusionment with theEWS, and a consequent loss of motivation and de-clining engagement, is to be avoided.

Building trust and establishing relevance may betime consuming and difficult, but these issues alsoencapsulate the promise of actively involvingcivilians in an EWS. All too often when monitor-ing ceasefires or conflicts, a political process that

represents the interests of the warring factionsand some notion of ‘balance’ between them is theprimary concern of the monitoring mechanism;further, mechanisms that do not directly involvecivilians may be paternalistic and dismissive ofcivilian concerns. By putting civilians at the heartof a monitoring mechanism, their concerns andinterests are foregrounded and must be consid-ered if the mechanism itself is to be effective. Thepolitical considerations that inevitably influencehow these mechanisms operate, in South Sudanand elsewhere, must be reoriented to includerather than exclude or minimize civilian con-cerns. If they do not, they risk the erosion of trustand credibility and the possible collapse of theearly warning system and the national and inter-national political fallout that would have.

The promise of civilian-led monitoring systems isnot that they are an alternative to ‘boots on theground’ or ‘monitoring lite’ – a ‘cheap’ or ‘easy’ al-ternative to a monitoring mission led by a team ofinternational experts. It is that instead of exclud-ing civilians and civilian perspectives and inter-ests from the political processes in which the EWSis embedded, it puts them front and centre. Thiscould arguably be a key to resolving a conflict ina way that serves the interests of the populationas a whole, rather than the interests of a narrowcircle of warlords.

This is also the advantage of having a civilian-ledEWS run by a CSO or even a coalition of civil as-sociations. If the mechanism is run by an interna-tional organization, associated with the UN or anyother multilateral body or for that matter a na-tional institution, civilian concerns will inevitablybe balanced against political or organizational im-peratives, whereas the imperative of a CSO will beto maintain the support of the civilian monitorson which the network depends, actively promot-ing the civilian perspectives they represent.


3Conclusion: Thepromise and challengeof civilian-led earlywarning systems

CEPO and Ceasefire have built and operated a successfulpilot fourth-generation EWS in South Sudan that is bothcivilian-led and civilian-focused. The experience hasproven the concept is viable and has significant potential,though it is not suited for conflicts in heavily militarizedareas in which civilians have been evacuated or had thechance to evacuate, or perhaps were never present in thefirst place. However, it is well-suited for conflicts fought inareas populated by civilians or in which civilians are ac-tively targeted or otherwise harmed, which is to say it issuited for most conflicts in the modern world.

Several key points have emerged from the CEPO/Ceasefirepilot. First and most critical is that each project must be tai-lored to the context in which it operates. This includes thenature of the conflict itself; the activists, civic associationsand CSOs that can be inducted into or constructed to formthe network; and the political environment in which theEWS operates. The key to a successful civilian-led EWS is inits adaptability to the conditions in which it operates.

Of fundamental importance to a civilian-led EWS is an ef-fective advocacy strategy. Without it, the EWS becomes anexercise in data collection. It needs an advocacy strategythat embodies the hope that their work will contribute toimproving their lives or the lives of others, otherwise civil-ian monitors will not have a reason to invest their time andenergy in it or justify the potential risk of their involvement.

The promise of civilian-led early warning systems is thatthey will of necessity foreground the interests and perspec-

tives of civilians in the political processes in which they areembedded. They should not be viewed as either an easy ora cheap alternative to mechanisms led by a team of inter-national experts; they require time and effort to constructand maintain, and the team that runs the EWS has to be asactive advocate for civilian interests in whatever politicalprocess it is embedded.

Civilian interests and perspectives are too easily marginalizedby peace agreements and political processes that privilege theinterests of the parties to conflict and/or the elites. Peace pro-cesses, or for that matter, inter-communal dispute resolutionprocesses that exclude civilian interests, can be paternalisticor dismissive of valid civilian concerns and as such may actu-ally be less effective. Civilian-led early warning systems are astep toward peace processes that are more inclusive of the in-terests and perspectives of the civilians they claim to serve.

RecommendationsFor early warning systems globally• Civilian-led early warning systems should be employed

when they are suited to the modalities of a conflict,that is, when the conflict is conducted in areaspopulated by civilians, when civilians are activelytargeted, or when they are otherwise subject to harm.

• Civilian-led early warning systems should be viewedas a way of incorporating civilian priorities andperspectives into decision-making processes relatingto conflict resolution.


• When designing conflict resolutionprocesses, whether these are formal peaceprocesses or mediation mechanisms to endinter-communal violence, policy makersshould recognize that civilian-led earlywarning systems require a seriouscommitment of time and resources. They arenot an afterthought, or a cheap and easyway of doing monitoring. They requiresignificant political and material support,and protection.

• Decision makers in conflict resolutionprocesses should be receptive toincorporating and prioritizing civilianperspectives in their decision-makingprocesses. They should recognize that thismakes decision-making processes moreeffective, and more likely to produce viableand durable resolutions to the conflict.

For early warning in South Sudan• Key national and international decision

makers must prioritize the protection ofcivilians in the ongoing violence in SouthSudan and support the inclusion of civilianperspectives into the peace process.

• Decision makers should recognize thepotential contribution of the EWS to thepeace process and to ending inter-communalviolence, and actively facilitate thatcontribution.

• International decision makers shouldsupport a national roll-out of theCEPO/Ceasefire civilian-led EWS, andcontinue to support it politically.

• The international community should be opento closer coordination with the EWS andmore responsive to its alerts to assistcivilians in need of protection.


1 Nyheim, D., Preventing Violence, War and StateCollapse: The Future of Conflict Early Warning andResponse, Paris, OECD, 2009, p. 22.

2 Palli, S., Early Warning Systems as a ConflictPrevention Tool: Recommendations for the ArabRegion, Beirut, Economic and Social Commissionfor Western Asia (ESCWA), E/ESCWA/ECRI/2017/WP.3.

3 Rupesinghe, K., FCE Citizen-based Early Warningand Early Response System: A New Tool for CivilSociety to Prevent Violent Conflict, Colombo, SriLanka, Foundation for Co-Existence, 2009, p. 7.

4 Marshall, M.G., Fragility, Instability, and theFailure of States Assessing Sources of Systemic Risk,Working Paper, New York, Center for PreventiveAction, 2008, p. 10.

5 Fragile States Index, Indicators:https://fragilestatesindex.org/indicators/p1/

6 Nyheim, D., Early Warning and Response to ViolentConflict: Time for a Rethink?, London, Saferworld,2015, p. 8.

7 Palli, op. cit., 2017, citing L. Gordenker, ‘Earlywarning: conceptual and practical issues’, in K.Rupesinghe and M. Kuroda (eds), Early Warningand Conflict Resolution, New York, St. Martin’sPress, 1989.

8 Nyheim, Preventing Violence, op. cit., p. 30.9 Global Partnership for the Prevention of Armed

Conflict, Early Warning and Early Response:Conceptual and Empirical Dilemmas, Issue Paper1, The Hague, September 2006, p. 18.

10 Ibid.11 UN Charter, chapters VI and VII,

https://www.un.org/en/about-us/un-charter 12 Ramcharan, B.G., ‘Preventive diplomacy at the

United Nations’, UN Chronicle, vol. 48, no. 2,2011.

13 Kriesberg, L., ‘The development of the conflictresolution field’, in I. William Zartman and J.Lewis Rasmussen (eds) Peacemaking inInternational Conflict: Methods and Techniques,Washington, DC, United States Institute of PeacePress, 1997.

14 Rupesinghe, op. cit., 2009, p. 8.15 Rummel, R.J., ‘Appendix II, Event data: bases of

empirical conflict analysis’, in R.J. Rummel,Understanding Conflict and War, vol. 4: War,Power, Peace, Beverly Hills, CA, Sage, 1979.

16 United States Agency for InternationalDevelopment (USAID), Famine Early WarningSystem Network, https://www.usaid.gov/results-and-data/data-resources/fews-net, accessed 27March 2021.

17 Palli, op. cit., 2017.

18 UNHCR, ‘Risk analysis and monitoring – refugeeemergencies’, Emergency Handbook. Geneva,UNHCR.

19 UN, An Agenda for Peace: Preventive Diplomacy,Peacemaking and Peace-keeping, Report of theSecretary-General, 17 June 1992, A/47/277 –S/24111, para. 23.

20 R2P is made up of three pillars: Pillar 1establishes state responsibility to ensure thesafety and security of its civilians; Pillar 2stresses the responsibility of the internationalcommunity to support states in this aim; andPillar 3 states that the international communityhas the obligation to ensure the protection ofcivilians when a state has manifestly failed to doso or when a state has targeted and attacked itsown citizens.

21 UN General Assembly, 2005 World SummitOutcome, October 2005, A/RES/60/1.

22 Ibid., paras 138–40.23 UN Office on Genocide Prevention and the

Responsibility to Protect, Early Warning,https://www.un.org/en/genocideprevention/early-warning.shtml, accessed 27 March 2021.

24 UN Office on Genocide Prevention and theResponsibility to Protect, Framework of Analysisfor Atrocity Crimes: A Tool for Prevention, NewYork, 2014.

25 See Genocide Watch and the Sentinel Project. 26 Stanton, G.H., 10 Stages of Genocide, Genocide

Watch, 2016, http://genocidewatch.net/genocide-2/8-stages-of-genocide/, accessed 27March 2021.

27 Nyheim, Early Warning, op. cit., p. 3.28 Palli, 2017, op. cit., p. 12.29 Rupesinghe, op. cit., 2009, p. 9.30 Ibid. See also Meier, P., ‘Sri Lanka: Citizen-based

early warning and response’, Conflict EarlyWarning and Early Response Blog, 5 March 2009.

31 Rupesinghe, op. cit., 2009, p. 8.32 Political Instability Task Force Homepage,

archived version, https://web.archive.org/web/20061208000556/http://globalpolicy.gmu.edu/pitf/, accessed 27 March 2021.

33 Wulf, H. and Debiel, T., Conflict Early Warning andResponse Mechanisms: Tools for Enhancing theEffectiveness of Regional Organisations? AComparative Study of the AU, ECOWAS, IGAD,ASEAN/ARF AND PIF, Crisis States Working PapersSeries No. 2, London, LSE, 2009.

34 The Centre for Systemic Peace Homepage statesthat financial support was terminated on 29February 2020, https://www.systemicpeace.org/

Endnotes


35 Rupesinghe, op. cit., 2009, p. 9.36 Wulf and Debiel, op. cit., 2009, p. 8.37 Ramsbotham, O., T. Woodhouse and H. Miall

(2005) Contemporary Conflict Resolution: ThePrevention, Management and Transformation ofDeadly Conflicts, 2nd edn, Cambridge, Polity, p. 113.

38 Rupesinghe, op. cit., 2009, p. 8.39 Nyheim, Preventing Violence, op. cit., pp. 3–4.40 International Crisis Group, ‘How We Work,

Methodology’, https://www.crisisgroup.org/how-we-work

41 See, for example, the webpage Crisis Watch2020, which provides recommendations to theEU on possible actions it can take in tencountries and regions at risk of conflict orescalation of violence, the majority of which areAfrican countries and regions, https://www.crisisgroup.org/crisiswatch/january-2021-alerts-and-december-trends-2020, accessed 28 March 2021.

42 Ramsbotham et al., op. cit., 2005, p. 114;Rupesinghe, op. cit., 2009, p. 10.

43 Rupesinghe, op. cit., 2009, pp. 10–11.44 Ibid.45 UNDP and OAS, Practical Guide: Early Warning

and Response Systems Design for Social Conflicts,Panama, UNDP, 2016, p. 15.

46 Ibid.47 Rupesinghe, op. cit., 2009, p. 11.48 Information on FCE is taken from Rupesinghe,

op. cit.., 2009, pp. 12–23.49 Senevirathna, P., ‘FCE’s early warning and early

response system’, chapter 4 in KumarRupesinghe (ed.) Third Generation Early Warning,Colombo, FCE, 2009

50 Nyheim, D., The Global Balance Sheet: EmergingSecurity Threats and Multilateral ResponseCapabilities, Stanley Foundation Working Paper,Muscatine, IA, October 2009, p. 7.

51 Meier, P., ‘Fourth-generation early warning andresponse (updated)’, Conflict Early Warning andResponse Blog, 6 March 2009.

52 Ibid.53 Alston, P. and C. Gillespie, ‘Global human rights

monitoring, new technologies, and the politicsof information’, European Journal of InternationalLaw, vol. 23, no. 4, 2012, pp. 1117–18, referred toin Ceasefire Centre for Civilian Rights, Eyes on theGround: Realizing the Potential of Civilian-ledMonitoring in Armed Conflict, London, Ceasefire,2017, p. 17.

54 Hersman, E., ‘The conflict early warningquandary’, Ushahidi Blog, 17 December 2008;Meier, P., ‘How Ushahidi can become a real earlywarning platform’, Conflict Early Warning andEarly Response Blog, 21 June 2009.

55 Doran, W., ‘How the Ushahidi platform works,and what comes next’, Ushahidi Blog, 5November 2018.

56 Meier, P., ‘New information technologies andtheir impact on the humanitarian sector’,International Review of the Red Cross, vol. 93, no.884, December 2011, p. 1239.

57 Ushahidi Staff, ‘Violence early warning systemfor Nigerian elections’, 27 March 2015.

58 Ceasefire Centre for Civilian Rights, op. cit., pp.18–19.

59 Meier, ‘Fourth-generation early warningsystems (Updated)’, op. cit.

60 Zenko, M. and R.R. Friedman, ‘UN early warningfor preventing conflict’, InternationalPeacekeeping, vol. 18, no. 1, 2011, p. 21.

61 Dorn, W., ‘Early and late warning by the UNSecretary-General of threats to peace: Article 99revisited’, chapter 14 in A. Schnabel and D.Carment (eds) Conflict Prevention from Rhetoric toReality, vol. 1, Lanham, MD: Lexington Books,2004, pp. 305–44.

62 Cockell, J., ‘Early warning and United Nationsreform’, RRN newsletter (Relief and RehabilitationNetwork), June 2003.

63 Ibid.64 Willmot, H., Improving UN Situational Analysis:

Enhancing the U.N.’s Ability to Prevent and Respondto Mass Human Suffering and to Ensure the Safetyand Security of Its Personnel, Stimson Centre,August 2017, p. 31, https:// www.stimson.org/wp-content/files/file-attachments/UNSituationalAwareness_FINAL_Web.pdf, accessed 28 March 2021.

65 Zenko and Friedman, op. cit.66 Ibid., pp. 22–3.67 UNDPPA, ‘What we do: Prevention and

mediation’, https://dppa.un.org/en/prevention-and-mediation

68 UN Peacekeeping, Protection of CiviliansMandate, peacekeeping.un.org/

69 UN Security Council Resolution 1996 (2011),S/RES/1996 (2011), para. 3(b)(ii).

70 Di Razza, N., Reframing the Protection of CiviliansParadigm for UN Peace Operations, New York,International Peace Institute, 2017, p. 3.

71 Ibid., pp. 3–4.72 MONUSCO, ‘Civil affairs’, https://monusco

.unmissions.org/en/civil-affairs73 Lazicky, C., Improving Conflict Early Warning

Systems for United Nations Peacekeeping, MAthesis, Harvard University, 2017, p. 5.

74 MONUSCO, op. cit.75 Lazicky, op. cit., p. 6.76 Ibid.77 MONUSCO, op. cit.


78 UN Peacekeeping, ‘Denied access to conflict-affected areas around Wau aggravateshumanitarian situation of population’, 27 August2018.

79 Stanton, G., ‘The Rwandan genocide: why earlywarning failed’, Journal of African Conflicts andPeace Studies, vol. 1, no. 2, 2009, pp. 6–25.

80 UN/SIDR, Developing Early Warning Systems: AChecklist, EWC III Third International Conferenceon Early Warning: From concept to action, Bonn,UN/ISDR, 2006, p. 2.

81 Global Public Policy Institute, ‘Making the UnitedNations system more effective on conflictprevention’, article, 24 July 2017, Berlin.

82 Ibid.83 Ibid.84 UN News, ‘“Human Rights up Front” initiative

ever more vital to strengthen UN’s preventivework, says deputy UN chief’, 28 January 2016,https://news.un.org/en/story/2016/01/521092-human-rights-front-initiative-ever-more-vital-strengthen-uns-preventive-work, accessed 28March 2021.

85 Damplo, D. and R. Saad, Policy or Aspiration:Shedding Light on the Current Status of the UN’sHuman Rights Up Front Initiative, New York,Universal Rights Group, 2019, p. 4.

86 Ibid., p. 6.87 Willmot, op. cit., p. 26; Zenko and Friedman, op.

cit.88 For example, just some of the early warning

systems sometimes cited but no longer inoperations are FAST, FEWER, SIPRI, UPPSALA andPIOOM.

89 Amao, O.B., D. Ettang, U. Okeke-Uzodike and C.Tugizamana, ‘Revisiting the utility of the earlywarning and early response mechanisms inAfrica: any role for civil society?’, Peace andConflict Review, vol. 8, no. 1, 2014, p. 81; Barrs, C.,‘Conflict early warning: warning who?’, Journalof Humanitarian Assistance, 12 February 2006,https://sites.tufts.edu/jha/archives/41

90 Nyheim, The Global Balance Sheet, op. cit., p. 7.91 Ibid., p. 4.92 Desmidt, S., ‘Conflict management and

prevention under the African Peace and SecurityArchitecture (APSA) of the African Union’, AfricaJournal of Management, vol. 5, no. 1, 2019, p. 82.

93 African Union, Constitutive Act of the AfricanUnion, 2000, Article 4(h).

94 Bah, A.S., E. Choge-Nyangoro, S. Dersso, B.Mofya and T. Murithi. The African Peace andSecurity Architecture: A Handbook, Bonn, FriedrichEbert Stiftung, 2014.

95 African Union Peace and Security, ‘TheContinental Early Warning System (CEWS)’, 15May 2018, http://www.peaceau.org/en/page/28-continental-early-warning-system-cews

96 African Union, Protocol Relating to theEstablishment of the Peace and Security Councilof the African Union, 2002, Article 12(3).

97 African Union Peace and Security, ‘TheContinental Early Warning System (CEWS)’, op.cit.

98 African Union, Continental Early Warning System:The CEWS Handbook, 7th draft, 2008, p. 17,https://www.peaceau.org/uploads/cews-handook-en.pdf

99 Sega, T.D., Conflict Early Warning Systems: ItsChallenges and the Continental Early WarningSystem, MA report, University of Witwatersrand,2017, p. 82.

100 African Union, CEWS Handbook, op. cit., p. 45.101 Ibid., p. 50.102 Desmidt, op. cit., p. 81.103 Institute for Security Studies, ‘The PSC should do

more to respond to conflicts in 2020’, pressrelease, 13 December 2019, https://reliefweb.int/report/world/psc-should-do-more-respond-conflicts-2020

104 The Member States of ECOWAS are Benin,Burkina Faso, Cabo Verde, Côte d’Ivoire, TheGambia, Ghana, Guinea, Guinea-Bissau, Liberia,Mali, Niger, Nigeria, Senegal, Sierra Leone andTogo.

105 Wulf and Debiel, op. cit., p. 17; Tejpar, J. andAdriana Lins de Albuquerque, Challenges toPeace and Security in West Africa: The Role ofECOWAS, Stockholm, Swedish Defence ResearchAgency, 2015.

106 ‘Use of early warning systems in Western Africato combat terrorism’, 2019, https://capstone.extension.harvard.edu/files/capstone/files/use_of_early_warning_systems_spring_2018_redacted.pdf; Amao et al., op. cit., p. 83.

107 Ibid.108 ECOWAS, ‘ECOWAS strengthens capacity of its

Early Warning Directorate on governance andhuman rights’, news item, 12 July 2017.

109 Wulf and Debiel, op. cit., p. 17.110 ‘Use of early warning systems in Western Africa

to combat terrorism’, op. cit., p. 4.111 WANEP, National Human Security Early Warning

Brief of GHANAWARN, quarter ending March2014; Information Systems in Africa, NationalEarly Warning and Response System, 2009,http://isiafrica.net/national-early-warning-and-response-system/, accessed 28 March 2021.


112 Hartmann, C. and K. Striebinger, ‘Writing thescript? ECOWAS’ military interventionmechanism’, in T. Börzel and V. van Hüllen (eds)Governance Transfer by Regional Organizations:Patching Together a Global Script, Basingstoke:Palgrave Macmillan, 2015, p. 75.

113 ECOWAS, Protocol Relating to the Mechanismfor Conflict Prevention, Management,Resolution, Peace-keeping and Security, adopted10 December 1999, Article 25.

114 Nyheim, Early Warning, op. cit., p. 15.115 Ibid. 116 Ibid.117 Sega, op. cit., p. 38.118 ‘ECOWAS at 40: Achievements and challenges’,

Statement by Major-General O.B. Akwa,Commandant, Kofi Annan InternationalPeacekeeping Training Centre (KAIPTC), 18 May2015, https://wanep.org/wanep/attachments/article/770/ecowas_at_40_commandant_statement.pdf, accessed 16 March 2021.

119 Ibid.120 IGAD comprises the countries of Djibouti,

Eritrea, Ethiopia, Kenya, Somalia, South Sudan,Sudan and Uganda.

121 Conflict Early Warning and Response Mechanism(CEWARN), IOM, https://micicinitiative.iom.int/micicinitiative/conflict-early-warning-and-response-mechanism-cewarn, accessed 28March 2021.

122 Nyheim, The Global Balance Sheet, op. cit.,October 2009, p. 6.

123 Ibid.124 Nyheim, Early Warning, op. cit., p. 10.125 For an infographic outlining the structure of

CEWARN see slide 5 of IGAD, Conflict EarlyWarning and Response Mechanism (CEWARN)Brochure, op. cit.

126 Saferworld, Towards a More Effective EarlyWarning System in the Horn of Africa: LearningLessons and Seizing Opportunities, London,Saferworld, 2015, p. 2.

127 Hassan, R., ‘CEWARN’s new strategy framework’,African Security Review, vol. 22, no. 2,2013, pp.26–38.

128 Saferworld, op. cit., p. 5.129 Ibid., p. 6.130 Ibid., pp. 16–17.131 See, for example, Leach, S., Preventing Violence:

Community-based Approaches to Early Warningand Early Response, Center for Security Studies(CSS), Zurich, ETH Zurich, 2016, p. 33.

132 Saferworld, op. cit., p. 17.133 Leach, op. cit.

134 Johnson, D.H., ‘Briefing: the crisis in SouthSudan’, African Affairs, vol. 113, no. 451, 2014, pp.300–9.

135 Thiong, D.A., ‘How the politics of fear generatedchaos in South Sudan’, African Affairs, vol. 117,no. 469, 2018, p. 624.

136 UN News, ‘Starvation used as weapon of war inSouth Sudan conflict, UN rights body finds’, 6October 2020, https://news.un.org/en/story/2020/10/1074742

137 Reuters Staff, ‘Study estimates 190,000 peoplekilled in South Sudan’s civil war’, 26 September2018.

138 UNHCR, ‘Refugees and asylum-seekers fromSouth Sudan’, Operational Data Portal: refugeesituations, data2.unhcr.org/en/situations/southsudan#_ga=2.75831317.1096969575.1602503666-180906835.1602503666, accessed 12October 2020.

139 Internal Displacement Monitoring Centre(IDMC), ‘South Sudan’, https://www.internal-displacement.org/countries/south-sudan,accessed 12 October 2020.

140 UNMISS ‘Civilians deliberately and brutallytargeted during surge in conflict in CentralEquatoria’, Press release, 3 July 2019.

141 UN Security Council Resolution 1996, Reports ofthe Secretary-General on the Sudan, adopted 8July 2011.

142 UNMISS, ‘Protecting civilians, monitoring humanrights and supporting implementation ofcessation of hostilities agreement’, Fact Sheet.

143 Stern, J., Establishing Safety and Security atProtection of Civilians Sites: Lessons from theUnited Nations Peacekeeping Mission in SouthSudan, Stimson Civilians in Conflict Policy Briefno. 2, September 2015, p. 6; UNMISS, ‘POCupdate’, 12–18 June 2020, https://unmiss.unmissions.org/sites/default/files/poc-update_12_-_18_june_2020.pdf, accessed 28 March 2021; UNPeacekeeping, ‘UN Protection of Civilians sitesbegin transitioning to conventionaldisplacement camps’, News item, 4 September2020.

144 Verjee, A., Ceasefire Monitoring in South Sudan2014–2019: A Very Ugly Mission, Peaceworks no.150, Washington, DC, United States Institute ofPeace (USIP), August 2019, p. 21.

145 South Sudan Peace Monitoring, ‘Background’,2020, https://ctsamvm.org/about-ctsamvm/,accessed 18 October 2020.

146 South Sudan Peace Monitoring, ‘Compositionand functions’,https://ctsamvm.org/composition-functions/,accessed 18 October 2020.

147 South Sudan Peace Monitoring, ‘Corecommitments’, https://ctsamvm.org/core-commitments/, accessed 18 October 2020.

148 Verjee, op. cit., pp. 21, 24–5. 149 Ibid., pp. 22–3. 150 Njoroge, H.E. Amb. Lt. Gen. A.S.K. (Rtd), Interim

Chairperson of RJMEC, On the Status ofImplementation of the Revitalised Agreement onthe Resolution of the Conflict in the Republic ofSouth Sudan, 1 January to 31 March 2019, Report002/19, para. 17, https://reliefweb.int/sites/reliefweb.int/files/resources/JMEC-1st-Qtr-2019-Report.pdf, accessed 22 October 2020, and Onthe Status of Implementation of the RevitalisedAgreement on the Resolution of the Conflict in theRepublic of South Sudan, 1 April to 30 June 2019,Report 007/20, para. 57, https://reliefweb.int/sites/reliefweb.int/ files/resources/RJMEC-2nd-Qtr-2020-Report-Final.pdf, accessed 22 October 2020.

151 Verjee, op. cit., pp. 13–15, 24–5. 152 De Waal, A., ‘When kleptocracy becomes

insolvent: brute causes of the civil war in SouthSudan’, African Affairs, vol. 113, no. 452, July 2014,pp. 361–2.

153 CEPO, ‘Programs’, http://cepo-southsudan.org/Programs, accessed 14 March 2021.

154 CEPO, ‘Mapping conflict violence in SouthSudan’, http://cepo-southsudan.org/mapping-conflict-violence-in-south-sudan, accessed 24February 2021.

155 CEPO, ‘Inter-communal violence requires urgentstate response’, Radio Tamazuj, 5 January 2021,https://radiotamazuj.org/en/news/article/inter-communal-violence-requires-urgent-state-response-cepo, accessed 7 January 2021.

156 Radio Tamazuj, ‘Kiir directs security apparatus toend armed conflicts, road ambushes’, 6 January2021, from Radio Tamazuj, https://radiotamazuj.org/en/news/article/kiir-directs-security-apparatus-to-end-armed-conflicts-road-ambushes, accessed 7 January 2021.

157 Thiong, op. cit., pp. 627–8.158 See: https://en.wikipedia.org/wiki/OODA_loop



Civilian-led early warning of violence in armed conflict is a viabletechnique that has significant potential to save lives by alertingboth responsible authorities and civilian populations themselves ofimpending threats. It is well-suited to the majority of today’s con-flicts, which frequently occur in populated areas and in which civil-ians are actively targeted or otherwise subject to serious harm.

Civilian-led early warning systems have the potential to: • Ensure that information on a wide spectrum of violence or

threatened violence is captured, including in particular sexualand gender-based violence, extortion, pillage and other exac-tions against local civilian populations, as well as communal vi-olence and patterns of criminal violence linked to the conflict;

• Provide real-time information from territories to which accessis difficult of denied for official monitors;

• Ensure that warnings quickly reach the communities that aremost under threat;

• Enable local community interests and perspectives to be reg-ularly inputted into peace processes;

• Recognise the active role civilians play in improving civilianprotection and security.

Informed by a detailed consideration of the historical develop-ment of early warning systems for various forms of political vio-lence, this report considers aspects of the role of civilians and civilsociety organizations in early warning systems in zones of conflict,and summarizes the lessons learned from the civilian-led andcivilian-focused early warning project run by the Community Em-powerment for Progress Organisation (CEPO) and the CeasefireCentre for Civilian Rights in South Sudan from 2019 to 2021. Thereport examines the pilot early warning system in South Sudan indetail, considering the context in which it operates, the challengesit has faced and the opportunities it has realized. It focuses on thepolitical considerations of operating in such a sensitive situationand the security concerns for staff and monitors that need to beaddressed in the design and implementation of the project.

Two key points that have emerged from this pilot project are first,that a civilian-led early warning system must be tailored to the

context in which it operates – to the nature of the conflict itselfbut also its social and political context – and, second, that withoutan effective advocacy strategy to protect civilians and advancetheir interests it becomes an exercise in data collection.

The promise embodied by civilian-led early warning systems isthat they will, of necessity, foreground the interests and perspec-tives of civilians in the political processes in which they are em-bedded. They should not be viewed as either an easy or a cheapalternative to mechanisms led by a team of international experts;they require time and effort to construct and maintain, and theteam that runs them has to be an active advocate for civilian inter-ests, which are too easily marginalized by peace agreements andpolitical processes that privilege the interests of the parties to aconflict and/or political elites.

Civilian-led early warning systems are a step towards peace pro-cesses that are more inclusive of the interests and perspectives ofthe civilians they claim to serve.

This report recommends that:• Civilian-led early warning systems should be employed when

they are suited to the modalities of a conflict, that is when theconflict is conducted in areas populated by civilians or whencivilians are harmed.

• Civilian-led early warning systems should be viewed as a wayof incorporating civilian priorities and perspectives into deci-sion-making processes relating to conflict prevention, de-es-calation and resolution.

• Policy makers should recognize that civilian-led early warningsystems require a serious commitment of time and re-sources. They should not be an afterthought, or a viewed ascheap and easy way of doing monitoring. They require signifi-cant political and material support, and protection.

• Key national and international decision makers in peace pro-cesses should support civilian-led early warning systems andrecognize that they offer a means of prioritizing the protec-tion of civilians and supporting the inclusion of civilian per-spectives into those processes.

In brief

This report has been produced with the financial assistance of the UK government. The contents of this report are the soleresponsibility of the publishers and can under no circumstances be regarded as reflecting the position of the UK government.

Ceasefire Centre for Civilian Rights3 Whitehall Court, London SW1A 2EL, United Kingdomwww.ceasefire.org @ceasefirecentre

first to know

Documents