CEWARN’s Strategy Framework and Civil Society Partnership: Lessons for early-warning and response systems

Introduction

After focusing on pastoralist conflicts in three cross-border areas for ten years, the Conflict Early Warning and Response Network (CEWARN) adopted an ambitious new Strategy Framework in 2012. The strategy transformed CEWARN into a system to detect and respond timeously to the potential escalation of violent conflicts that may be triggered by economic, environmental, governance, security and social matters in all member states of the Intergovernmental Authority on Development (IGAD).¹

Besides expanding the thematic and geographical range, the strategy enhanced CEWARN’s partnership with civil society, which was initiated by the pioneering CEWARN Protocol in 2002.²

This article reviews the implementation of CEWARN’s partner-ships with civil society organisations (CSOs) for data collection, analysis and response under the current Strategy Framework. It identifies outstanding challenges and lessons for (inter)governmental conflict early-warning and response systems (CEWRS).

The Netherland’s Partners for Resilience, a civil society organisation, established early-warning systems for drought, flood and conflict in Africa. Photo: Raimond Duijsens/NLRC

The experiences of CEWARN and other (inter)-governmental early-warning (and response) systems show that the participation of CSOs, including local community-based organisations, expert non-governmental organisations (NGOs) and research institutions, can prove beneficial at all stages of the early-warning and response process, including data collection, analysis, warning and the implementation of responsive action. Civil society reporters may collect data on local-level dynamics that cannot be captured through media monitoring, highlight human security concerns, and assist in the verification and evaluation of incident reports.³ Research institutions may enrich analyses by adding alternative perspectives to those of (inter)-governmental officials and proposing complementary responses.⁴ In local-level responses, community leaders and CSOs may warn local stakeholders, undertake rapid preventive interventions, and facilitate dialogue among community members.⁵ In high-level crisis responses, specialist conflict resolution NGOs may provide technical support to mediators. In long-term preventive measures, non-governmental experts may assist in policy development to tackle countries’ structural vulnerabilities.⁶ Participation is hoped to augment the legitimacy and acceptance of conflict interventions of intergovernmental organisations (IGOs).⁷ To reap these benefits, governments and CSOs must navigate organisational, technical and political challenges.⁸

Besides IGAD, the African Union (AU) and six Regional Economic Communities (RECs) have established early-warning (and response) systems or relevant capabilities.⁹ These systems differ in their design and priorities. However, several utilise technology that was created for the CEWARN Reporter, feature national centres analogous to CEWARN, and work with CSOs for data collection, analyses and responses. CEWARN’s 20 years of experience in partnering with civil society, therefore, holds lessons for CEWRS that grapple with similar challenges.¹⁰

This article reviews the implementation of CEWARN’s civil society partnerships under the Strategy Framework to identify challenges and draw lessons for (inter)governmental CEWRS and CSOs. To this end, it examines how CSOs contributed to the collection of early-warning data, analyses, and responses since the adoption of the Strategy Framework. The study is based on 20 key information interviews with IGAD and CSO representatives.¹¹

CEWARN – IGAD’s Conflict Early Warning and Response Mechanism – was established in 2002 on the basis of a protocol signed by IGAD Member States.

The study shows that the Strategy Framework enables civil society participation at all stages of the warning-response process, but the modalities of its implementation bear technical, operational and political challenges that require continuous experimentation. The findings show that:

Expanding the net of community monitors who reported local-level incidents in the original cross-border clusters across all IGAD countries was unfeasible. Therefore, CEWARN built networks of NGO reporters, whose primary role is to verify and evaluate data that is collected through media monitoring.
The integration of 45 National Research Institutes (NRI) with sectoral specialisations in seven nations enhanced CEWARN’s analytical capacity. But producing periodic countrywide sectoral analyses remains challenging for scarcely resourced institutes. Crucially, the sensitivity of sharing security-related information complicates research partnerships.
The multi-agency model of CEWARN’s Conflict Early Warning and Response Units (CEWERU), in principle, ensures the coordination of government and CSOs in planning and implementing preventive action. The model, however, hinges on the full operationalisation of CEWERU Committees by states.
Consolidating a funding model for NGOs and NRIs to mobilise third-party funding to contribute to CEWARN is the crux to sustaining the envisaged partnerships.
Continuously building confidence between government and civil society, and maintaining sufficient civic space for CSOs to undertake early warning and response activities are essential to implement the Strategy Framework and envisaged partnerships.¹²

Initially CEWARN focused on pastoralist conflicts in three border areas between Ethiopia and her neighbours, Djibouti, Somalia, Kenya, Uganda and South Sudan. Photo: Rod Waddington

CEWARN’s original cross-border clusters and evolving architecture

CEWARN was designed as a decentralised system to enable early warning and responses by IGAD states on the national and local level. The CEWARN Unit in Addis Ababa is the linchpin of the system that became operational in 2003. It coordinates data collection, analyses, information-sharing, and joint responses between governments. CEWERUs that are embedded in ministries oversee all operations in countries and orchestrate responses.¹³

For ten years, CEWARN focused on pastoralist conflicts in three border areas between Ethiopia and her neighbours, Djibouti, Somalia, Kenya, Uganda and South Sudan.¹⁴ Whereas borderlands are sensitive to national security, the rationale for initiating CEWARN in the Dikhil, Karamoja and Somali cross-border clusters was to address conflicts that required cooperation between governments and gain their confidence in CEWARN.¹⁵ Intrastate conflicts, meanwhile, fell outside CEWARN’s mandate. To prevent pastoralist conflicts, CEWARN established a network of field monitors and peace committees in the clusters.¹⁶

The contributions of CSOs to CEWARN’s operationalisation and original clusters are well-documented.¹⁷ While international NGOs were instrumental in designing CEWARN, a commercial software developer, Virtual Research Associates, created the CEWARN Reporter.¹⁸ Incident and situation reports, which individual community-based monitors entered into the CEWARN Reporter, constituted the primary data collection method and permitted CEWARN to develop predictive capability for violent incidents, such as cattle raids.¹⁹ To prevent incidents, CEWERUs alerted community leaders and law enforcement to intervene. Civil society actors were represented in local peace committees that effected preventive measures.²⁰ CEWARN’s Rapid Response Fund served to finance interventions and prevention initiatives.²¹

Studies on CEWARN clusters highlighted several challenges, including poor communication infra-structure in remote areas and biased reporting, training deficits, and information overload among community monitors.²² Analyses of CEWARN in Kenya and South Sudan mentioned distrust among community members and local officials’ vis-à-vis monitors.²³ CEWARN’s decentralised design was lauded for fostering democratic security governance by including CSOs in CEWERU Committees.²⁴ The operationalisation was, however, protracted, with one study finding that by 2013, three countries had functional CEWERUs.²⁵ Establishing CEWERUs proved time-consuming as their capabilities and location within ministries had to be negotiated. State officials were hesitant vis-à-vis the intergovernmental information-gathering system and involvement of non-state actors.²⁶ Since the CEWARN Protocol did not make inclusive CEWERU Committees compulsory, some governments solely appointed CEWERU heads.²⁷ Several studies pointed to CSOs’ dissatisfaction with the lack of access to collated CEWARN data and absence of their concerns in published CEWARN reports.²⁸

The strategy change of 2012 would transform CEWARN into a system to monitor conflict factors relating to economic, environmental, governance, social and security across the territories of all IGAD countries. The thematic expansion shifted the focus from rapid local responses to early warning and policy development on the IGAD and national level. Whereas the CEWARN Unit and CEWERUs remain the cornerstones, the infrastructure of cross-border clusters was merged into national architectures.CEWARN also adopted new data collection methods that are explained in Figure 1.²⁹

Figure 1: Civil society participation in the early-warning and response process in IGAD and member states

Policy framework

Civil society participation in CEWARN is envisaged in IGAD policies but the modalities are within governments’ discretion. The Agreement establishing the IGAD encourages collaboration with non-governmental agencies in pursuit of IGAD’s objectives which include promoting peace.³⁰ The CEWARN Protocol provides that national CEWERU Committees comprise representatives of the government, parliament, police, military and non-state actors including NGOs, religious organisations and research institutions. Governments shall promote the involvement of NGOs, academics and community groups in data collection. CEWERUs should conduct analyses with research institutes that are represented in Committees and with independent academic institutions. Analyses should be available to the CEWARN Unit and “to the greatest extent possible” to civil society, but states may restrict access to information on grounds of national security. To effect early responses, governments may involve CSOs in alerting wider society and resolving grassroots conflict.³¹

By expanding the thematic and geographic scope to prevent conflicts across IGAD countries, the Strategy Framework modified the role of civil society reporters, research institutes and responders. It underscores the mobilisation of government and civil society as CEWARN’s key strength but acknowledges a lack of analytical, communication and implementation capabilities among national actors. The strategy aspires to build relevant capabilities in CEWERU Committees and involve citizens and private enterprises to promote ownership and human security.³²

At its introduction, scholars greeted the ambitious strategy with scepticism, owing to the scarcity of resources and governments’ reluctance to accept warnings and address conflicts within their jurisdiction.³³ The strategy not only exceeded the Protocol in terms of the thematic breadth and CEWARN’s policy development function, but posed political challenges by tasking CEWARN to address governance-related vulnerabilities.³⁴ Since the Protocol was not amended, the modalities of participation in CEWERUs remain at the discretion of governments.³⁵

IGAD’s Regional Strategy likewise aims to involve CSOs in conflict prevention and in promoting good governance, democracy and human rights.³⁶

CEWARN partnered with established NGOs and academic institutions to recruit and train reporters. Photo: Global.finland.fi

The Regional Gender Strategy identifies CSOs and CEWARN as key implementing agencies to foster women’s participation in prevention and peacemaking.³⁷ A draft Protocol on Preventive Diplomacy and Mediation, which is not yet in effect, promotes civil society inclusion in mediations and foresees that CEWARN systematically supports IGAD peace diplomacy through analyses.³⁸ In sum, IGAD policies envisage civil society contributions at all stages of the warning-response process but let governments decide on the extent to which CSOs may participate.

Data collection

The implementation of the Strategy Framework began in 2013 but data collection on the five analytic domains started in 2016. The strategy made it necessary to expand data collection methods and external partnerships.³⁹ Besides government statistics for structural vulnerability analyses, CEWARN planned to leverage geospatial data and enable crowdsourcing by permitting observers to report incidents via SMS.⁴⁰ Automated media monitoring became a central data collection instrument, while the benefits of collecting information on a spectrum of issues across countries through a resource-intensive reporter network were weighed against the cost.⁴¹

Rather than solely relying on media sources, CEWARN decided to modify the algorithm and questionnaire of the CEWARN Reporter and to develop nationwide networks with a maximum of 60 reporters per country. In lieu of relying on individual community-based field monitors as in the initial clusters, CEWARN partnered with established NGOs and academic institutions to recruit and train reporters.⁴²

CEWARN retained and modified the reporter system for several reasons. Firstly, since “CEWARN is about democratising security,” IGAD committed to partnering with NGOs to foster public ownership of the system and ensure that it did not duplicate state intelligence. Secondly, expanding the net of field monitors, whose reach was limited to their immediate surroundings, to cover entire national territories was impractical. Therefore, CEWARN partnered with NGO officers and academics who could draw from their respective networks of informers, brought relevant knowledge on a wider geographical area, and were better equipped to assess the relevance of information. The better-qualified reporters were not only hoped to improve reporting but to balance out biases in media-based data.⁴³ Thirdly, as the monitoring network in the clusters cost US$ 0.6 million per annum, augmenting the number of community monitors to cover all IGAD countries was financially unsustainable. NGOs, in contrast, were expected to raise donor funding independently and were compensated minimally for their contribution to CEWARN.⁴⁴ Fourthly, whereas community members were more willing to share information with civil society than government actors, CEWARN abandoned the original model as it could not guarantee the safety of local monitors, who had been accused of “selling out” and targeted by violence-makers in their communities.⁴⁵

NGOs were selected owing to their ability to cover analytic domains and vetted in coordination with governments.⁴⁶ While CEWARN set technical standards, governments’ approval for the selection was indispensable to maintain trust in CEWARN data and analyses. The negotiated selection implied a bias against outspoken NGOs and compromises regarding NGOs’ capabilities.⁴⁷ Civil society representatives raised concerns that the rigorous selection encourages self-censorship among reporters, as does the strict legislative regulation of civic activities in IGAD countries.⁴⁸

By 2021, CEWARN featured networks with up to 60 trained NGO reporters in seven IGAD countries and continuously gathered information in all analytic domains through the CEWARN Reporter. Since NGOs were concentrated in urban centres, the network was thin in rural areas, and the NGO reporters could not complement the media monitoring data with information on local-level dynamics on a meaningful scale.⁴⁹ The quality of the initially promising data gradually deteriorated as fatigued reporters entered inaccurate information. The envisaged model eventually proved unsustainable owing to a “mismatch of expectations.” Whereas CEWARN offered training and equipment, NGOs expected to be fully compensated for reporting tasks rather than raise third-party funding to provide their services. CEWARN, meanwhile, could not acquire the required funds centrally.⁵⁰

Rather than abandoning civil society reporters, who were integral to the system from the onset, CEWARN is presently testing readjustments to the reporter model. Instead of complementing the wealth of information from media monitoring with reports on local-level incidents, the primary function of NGO reporters now consists of verifying, evaluating and contextualising media-based findings.⁵¹

The Head of the African Union Mission in Somalia, Ambassador Francisco Madeira (left) with IGAD Executive Secretary, Dr Workneh Gebeyehu in Mogadishu, Somalia (18 February 2020). Photo: AMISOM/Omar Abdisalan

Early-warning analysis

CEWARN analyses are based on the above-described data sources and conducted by country experts at the CEWARN Unit and analysts in CEWERUs, who are responsible for verification, triangulation and analysis.⁵² Country-specific information is only accessible to data managers, national analysts, the CEWARN directorate and CEWERU. The CEWARN Unit does include NGO-seconded analysts, but non-state actors contribute to analyses through NRIs and, in rare instances, CEWERU Committees.⁵³

While the CEWARN Protocol envisaged one NRI per country, the new strategy increased the number to five research institutes that specialise in an analytical domain bringing the total to 45. The NRIs comprise government-approved universities and think tanks that were chosen owing to their sectoral expertise with the intent to tap into their ongoing work. NRIs issue quarterly situation reports on developments in their thematic domain to the CEWARN Unit and CEWERU.⁵⁴

The expansion to 45 specialist NRIs has enhanced CEWARN’s analytical capacity. NRI reports are, according to CEWARN’s Director, valued by CEWERUs whose Committees include the NRIs to develop response options.⁵⁵ Challenges in implementing the research partnership, firstly, result from the limited capacity of scarcely resourced institutes to conduct country-wide sectoral analyses periodically.⁵⁶ Secondly, whereas CEWARN compensated NRIs in the past, their contributions are now expected to be financed by member states or donors.⁵⁷ Thirdly, although CEWERUs should grant NRIs selective access to CEWARN data,⁵⁸ data-sharing with the non-state entities remains tightly controlled owing to the sensitivity of security-related information.⁵⁹ Fourthly, since academic freedom is limited and fluctuates in IGAD countries, researchers must exercise caution when analysing governance-related conflict vulnerabilities. Finally, the system would require a continuous feedback mechanism to ensure NRI analyses are adequate for decision- and policymaking.⁶⁰

In Kenya, whose peace infrastructure had inspired CEWARN, CSOs other than NRIs could contribute to analyses and drive the agenda of the CEWERU Committee.⁶¹ This was, however, not feasible in CEWERUs of governments that deemed sharing security-relevant information for analyses too sensitive.⁶²

Somaliland’s biggest livestock market, Burao, Somalia. Photo: ILRI/Peter Ballantyne

Warning and response

Prior to CEWARN’s reform, the response component was the outstanding feature of the decentralised system and the main activity in cross-border clusters. Besides prompting rapid interventions by police and community leaders to mitigate violence, CEWARN developed and accompanied the implementation of prevention policies including a cross-border trade regime and livestock markets to prevent cattle rustling.⁶³

The new strategy shifted CEWARN’s focus from local-level responses to early-warning capabilities and policy-development on the national and IGAD level. When aiming to recommend responses to tackle vulnerabilities relating to economic, environmental, governance, social and security matters, CEWARN “bit more than we can chew” given the breadth of response options and responders. Whereas CEWARN now primarily functions as an early-warning system, it is reinvigorating its response component and strengthening its policy-development capabilities for structural conflict prevention.⁶⁴

CEWARN enables government responses and does not take independent action, but the CEWARN Unit informs IGAD decision-makers, who may prompt IGAD responses, including preventive diplomacy, mediation and policy-development in the IGAD Secretariat. Early-warning reports are evaluated by the CEWARN Director and, if required, submitted to the Peace and Security Division Director and IGAD Executive Secretary. The latter may decide to inform states’ Permanent Secretaries, the Council of Ministers and Assembly of Heads of State and Government.⁶⁵

CEWRS commonly grapple with a warning-response gap that is often attributed to the difficulty of communicating early-warning information and a lack of political will by decision-makers to take costly preventive action.⁶⁶ In CEWARN’s decentralised system, the gap relates to the fact that states are responsible for responses.⁶⁷ Communicating warnings to inform policy-making by governments is particularly challenging and a focal point of CEWARN’s communication strategy with Permanent Secretaries.⁶⁸ Within IGAD, the gap may be narrowed by streamlining standard operating procedures for the communication of warnings to avoid bottlenecks and ensure politically inconvenient warnings are passed on.⁶⁹

In IGAD, early-warning and mediation support are not formally integrated as no standard procedures exist to leverage CEWARN’s analytical capacity to plan mediations by Special Envoys. In practice, CEWARN produces analyses to support mediations when the need arises. CEWARN staff who are seconded to Special Envoys’ offices provide informal links between the early-warning and mediation structures, which are coordinated in meetings of the Peace and Security Department.⁷⁰

IGAD’s peace diplomacy provides few opportunities for civil society participation. The Mediation Support Unit (MSU) partners with NGOs for knowledge management and the mediation roster features civil society actors, but these structures are not used for mediations.⁷¹ IGAD-mediated dialogues to negotiate the original and Revitalised Agreement to Resolve the Conflict in South Sudan featured civil society delegations.⁷² Inclusion mechanisms, however, had little bearing on negotiation outcomes that were determined in closed meetings between leaders of conflict parties and IGAD states.⁷³

CEWRS may prompt both crisis interventions and long-term prevention.⁷⁴ CEWARN seeks to shift the modus operandi from issuing “fire alerts” to forecasting long-term trends to address structural vulnerabilities through policymaking.⁷⁵ The CEWARN Unit, therefore, conducts national and regional conflict profiling and scenario building, flags conflict drivers such as taxation to governments, and explores policy options.⁷⁶ While sensitive country analyses are confidential, CEWARN publishes regional reports that warn against conflict drivers including electoral mismanagement in general terms.⁷⁷ Preventive policy responses are not only hard to achieve because it is difficult to communicate abstract links between structural issues and violent conflict, but their timely implementation hinges on national legislative processes and bureaucracies.⁷⁸

National responses are coordinated by the CEWERUs that are located in ministries or offices of vice presidents. CEWERU heads have access to CEWARN’s country-specific data and may inform government agencies and CSOs to take action.⁷⁹ The Strategy Framework aspires to establish subnational CEWERUs and peace committees that, however, must yet be operationalised in most countries, and reinvigorated in Kenya.⁸⁰

The multi-agency model of CEWERU Committees, in principle, ensures that CSOs participate in the planning and execution of responses.⁸¹ While Kenya’s and Uganda’s CEWERU Committees regularly meet and involve CSOs, the capacity, activities and participation of CSOs in Ethiopia’s and other CEWERUs have declined owing to government restructuring and intrastate crises.⁸² The CEWARN Unit, therefore, promotes the revitalisation of CEWERUs and advocates statutory reporting procedures for Committee meetings and responsive action.⁸³ Whereas CSOs were integral to CEWARN responses in cross-border clusters, the same must yet be achieved under the Strategy Framework.

Conclusion

CEWARN’s 20 years of experience in partnering with CSOs and the Strategy Framework’s implementation hold lessons for inter(governmental) CEWRS, for CEWARN illustrates the benefits of civil society participation and solutions to pertinent technical, operational and political challenges.

Collecting data on local-level incidents through community monitors across national territories, CEWARN’s experience suggests, is not feasible in most contexts as it would require vast and costly networks of field monitors. NGO reporters, however, may contribute their knowledge to verify, evaluate and contextualise data that is gathered through media monitoring.⁸⁴

Thanks to their thematic and country expertise, research institutions can boost the analytical capacity of CEWRS, but scarcely resourced institutes in conflict-affected countries may lack the capacity to periodically produce nationwide sectoral analyses. The sensitivity of sharing conflict-related information between (inter)governmental and non-state agencies is an obstacle to research partnerships that require a sound knowledge management regime.⁸⁵

CEWARN’s decentralised design and multi-agency CEWERU Committees, in principle, enable the coordination of state organs and CSOs in the planning and implementation of crisis responses and preventive activities. Statutory reporting procedures may be required to ensure that national-level responses are, indeed, implemented and inclusive CEWERU Committees fully operational.⁸⁶

African (inter)governmental CEWRS’ partnerships with CSOs hinge on the latter’s ability to mobilise third-party funding from international development partners to finance their services. Although successful in West Africa, this model has proven unsustainable in CEWARN, where NGO networks that can reliability raise funding and feed into CEWRS must yet be built.⁸⁷ IGOs, CSOs and development partners must explore alternative funding models or ways to accelerate the development of sustainable NGO networks.

CEWARN is committed to “democratising security” by promoting civic participation in conflict prevention and a human security agenda.⁸⁸ IGAD countries committed to involve CSOs when approving the CEWARN Protocol and Strategy Framework, but remain hesitant to share information and include non-state actors in managing security matters in practice.⁸⁹ Whereas participating CSOs get a seat at the table in CEWERU Committees, a perception that CEWARN leverages CSOs to gather information for states persists among civil society actors.⁹⁰ Building confidence between governments and civil society and opening up civic space for CSOs to engage in conflict prevention are, therefore, essential to bringing the CEWARN strategy and its partnerships to fruition.

Michael Aeby (PhD) is a Research and Teaching Fellow at the University of Basel Centre for African Studies and the Institute for Democracy, Citizenship and Public Policy in Africa at the University of Cape Town.