Choosing the team

Team composition

Conflict analysis can be led by an internal team, or by external consultants. The conflict sensitivity literature recommends that conflict analysis be carried out by the commissioning actor’s in-house staff, to improve the impact of the analysis, to ensure the findings and recommendations are relevant to the commissioner, to challenge internal staff assumptions and develop capacity and to ensure ownership and that the paper is read, understood and internalised (e.g. Conflict Sensitivity Consortium, 2012: 5–6). However, there are many trade-offs and challenges to consider. For example, when time is limited, it can be more efficient to use specialised consultants with experience using the analytical tools and methodologies or with greater knowledge of the context. Different commissioners/publishers will have different capacities and strengths: INGOs may be better equipped to use participatory approaches; donors may have greater access to key informants and intelligence reports; donor involvement in interviews may raise sensitivities.

Incorporating and understanding the breadth of national and local perspectives are crucial to understand local realities and the latest conflict dynamics. Many of the tools are designed to be participatory, to include views from, for example, elites, communities, organisations and locally based staff. Careful actor analysis should select participants from across all groups, paying attention not to reinforce exclusive conflict dynamics and to ensure the inclusion of minorities, women and groups that might oppose peace (Conflict Sensitivity Consortium, 2012). Discussing conflict can be very sensitive and topics should be carefully chosen. Finding neutral places for dialogue, protected from external agendas, can move discussions forward (Peacebuilding Center, 2013). Local staff and communities are embedded in the context and may have less objectivity, making triangulation of findings important. Challenges include who to include, whose voices and interests to prioritise, the level and quality of participation and ownership of the processes (Schmelzle, 2005; Donais, 2009). Shahab Ahmed (2011) examines the use of PCIAs in Pakistan (2006–8) and finds a disjuncture between the use and importance given to PCIA and the realities of the staff in country. The paper found a lack of understanding among people working at the project and implementation levels of what the PCIA was, why it was needed and the theory behind it, and that PCIA and ‘conflict sensitivity’ were ‘alien concepts’ for the majority of staff (ibid.).

Perceptions and bias

Understanding implicit bias in the data

Most violence monitoring systems use media reports to track conflict events, therefore the data are dependent on the integrity, quality, capacity and reliability of the media reporting. Potential biases should be examined. Parks, Colletta & Oppenheim (2013) highlight that spatial bias can occur, as there are fewer media sources in rural areas and a predominance of urban reporters who might not know the rural areas. This problem is exacerbated when conflict occurs in rural areas – such as in Thailand’s Deep South and in Mindanao. Another potential bias lies in reporters’ interpretation of what happened, who was involved and what the motivations were (e.g. criminal violence or insurgency-related violence). ‘In areas such as Mindanao, where multiple forms of contestation overlap, this may lead to misdiagnosis of fundamental conflict dynamics and trends’ (pp. 93–4). The paper suggests mitigating against these forms of bias through triangulation of multiple data sources and combining data analysis with more detailed qualitative analysis.

Source: Parks et al. (2013).

 
Perceptions are central to understanding conflict, and conflict analysis should examine and understand the experience of conflict from a variety of perspectives (Fisher et al., 2000). Conflicts are born of disagreements, and there are often few truths. While the news media tends to provide information on ‘conflict behaviour, actors, events and issues’, it does not tend to cover ‘perceptions, reasoning, motivations and assessments of the parties to a conflict’ (Höglund & Öberg, 2011: 7-8).

Sources of information and data may be biased – explicitly or implicitly – and it is important to examine their rigour, accuracy and potential biases. In its evaluation, CDA Collaborative (2013: 4) found conflict analyses could be biased and narrow because people either work with implicit assumptions based on personal experiences or explicitly use conflict analyses to make a case for their desired action.

The conflict analysis process is useful in prompting staff to question their own assumptions and to consider gaps and differing interpretations of issues. The Conflict Sensitivity Consortium (2012: 6) highlights that ‘part of conflict sensitivity is recognising that project staff form part of their contexts and may interpret situations based on their own histories’. Training may be required for staff – both those carrying out the analysis and those using it. For example, conflict analyses are often gender blind. Anderlini (2006) suggests holding short-term training or workshop sessions on the significance of gender issues to conflict.

Another challenge is that, when policy/political priorities frame the focus of analysis, there is a risk of the analysis overemphasising donor priorities, which might not be the most important issues for local communities, for the country as a whole, or for securing peace. Yet integrating conflict analysis into strategic and policy processes makes it more likely to affect policy and practice. To address these challenges, conflict analyses typically draw on and commission a range of information sources – both independent from and embedded in policy processes.

The OECD (2012: 78) suggests the following questions to help manage implicit assumptions when choosing a conflict analysis tool:

  • Do the evaluators share the underlying assumptions about the conflict that form the basis for analysis?
  • Is the tool’s understanding/assumptions about the nature of conflict appropriate to the specific context in which the programme or policy is being implemented?
  • Does this perspective correspond to the mandate and values of the organisation being evaluated?