[This post is part of a series on re-thinking conflict early warning.]
When I met with them a year ago, the board of the Iraqi Centre for Negotiation and Conflict Management had pretty clear ideas about early warning. Any system they set up to monitor overt threats or violent incidents that could trigger a conflict would be too slow. Journalists would get to these events faster; the limited time of the Centre’s mediators would be best spent keeping up with the news. Even if they did manage to set up a near real-time incident and threat monitor, they argued that these events were so close to the outbreak of conflict that it was often too late for preventive action. When we discussed what could help them move from reactive mediation to proactive intervention, they honed in on collecting perceptions data. They wanted to know how people on the street were feeling, what the pulse of a community was. What they needed was a more systematic way to tap into this pulse.
The Mercy Corps Iraq team (who support the Centre) is still in the process of rolling out its perceptions monitoring system (see this blogpost for an overview). But other teams working with peacebuilders and peace activists in conflict and post-conflict contexts have similarly found that perceptions monitoring is the best approach to conflict early warning. From 2007 to 2012, UNDP Sudan’s Crisis and Recovery Mapping and Analysis (CRMA) project carried out community-level perceptions mapping in six states of Sudan and ten states of South Sudan, in collaboration with the respective State Governments. The CRMA methodology built on existing tools such as Rapid Rural Appraisals, Conflict Analysis Frameworks, Vulnerability Assessments and Community-Based Risk Assessments to provide an evidence-base generated at the grassroots. Community workshops were run by the CRMA team in partnership with Government officials to gather perceptions about threats and risks. Each workshop gathered about 30 participants, from mixed backgrounds representing the community, over the course of two days. The workshops ran a variety of exercises, including plenary fora, participatory mapping, mind mapping and focus groups. Community perceptions gathered at these workshops were then assigned a category and a geographic location, to allow for both thematic and geographic analysis.
Through its work, CRMA supported government and civil society actors to jointly identify priorities for intervention and response. The process fostered an open dialogue, strengthening the capacities of local actors to respond to potential conflicts in a timely and appropriate manner. For example, in South Kordofan, UNDP Sudan had carried out its community level mapping exercise in collaboration with the state’s Reconciliation and Peaceful Coexistence Mechanism (RPCM) – a body which (prior to the conflict that started in South Kordofan in 2011) was similar to the Iraqi Centre for Negotiation and Conflict Management. UNDP supported the RPCM to analyse perceptions data from the community level mapping exercise in order to inform its priorities for action. (I gave an Ignite talk about this work a few years ago.)
Analysis of the perceptions data collected by CRMA was always somewhat problematic: the perceptions are not collected from a random sample of people or places, making any statistical analysis impossible. At most, the analysis allowed for identifying some geographic and thematic patterns that mediators and planners should look into. In Cyprus, a joint initiative of SeeD and UNDP has developed a more complex model of interpreting perceptions for the benefit of peacebuilding programs. Their Social Cohesion and Reconciliation Index (SCORE) is made up of five pillars. Four of them collect background information on civil society actors, local governance, political parties and local media – they are important to the model, but less relevant to this discussion. The fifth pillar collects responses to a questionnaire on perceptions from a random sample of Cypriots. The questions provide numerical scores for 25 indicators of social cohesion and reconciliation, such as negative stereotypes, intergroup anxiety, ingroup identification, trust, shared vision, etc. Five of these indicators are further aggregated to give an overall social cohesion score; another five are aggregated to give an overall reconciliation score. Each questionnaire respondent is tagged by their location and demographic characteristics to allow for comparisons between geographic areas and between groups.
Using a combination of statistical techniques to explore this perceptions data, the SCORE team can provide useful insights for peacebuilding practitioners. The model identifies the factor contributions of each variable to the composite variables of social cohesion and reconciliation. For example, it finds that whether Greek Cypriots trust Turkish Cypriots are much stronger predictors of negative stereotypes than whether Greek Cypriots have been in contact with Turkish Cypriots. This suggests that peacebuilders should focus their efforts on confidence building measures over projects that increase contact. The model also shows correlations between any pair of indicators. For example, it finds that negative stereotypes and social anxiety have a strong positive correlation, suggesting that programs should discuss together the stereotypes groups hold and the threat they feel the other group poses to their culture and social structure. Combining these findings with observations about the different indicator values between geographic areas or between groups results in even more precise recommendations. For example, contact between Greek Cypriots and Turkish Cypriots has a much stronger effect on reconciliation in Limassol than it does in Nicosia, suggesting how peacebuilders should target contact programs.
The first SCORE questionnaire in Cyprus has been completed and preliminary analysis can be viewed in this prezi. The team is now developing an online tool that will allow the public to browse indicator scores and graphical representations of how indicators correlate with each other, helping them draw their own conclusions. Plans to roll out the index to other countries are also afoot.
Too often, early warning systems collect vast amounts of data whose analysis leaves actors on the ground none the wiser about what course of action to take in order to prevent conflict or respond to an evolving situation. Despite differences in complexity and approach, the CRMA and SCORE initiatives have in common their focus on answering the practical, action-oriented questions peacebuilders face. These questions are often about what people are thinking or feeling, the pulse of a community that the Iraqi mediators talked about. Focusing on ways to capture and analyse perceptions is one way to build conflict early warning systems that truly support preventive programming.