Choosing the method


Conflict analyses are typically funded, published and/or written by policymakers, practitioners (especially INGOs), think tanks, policy-oriented research centres, human rights organisations and private sector actors / consultancies. The conflict sensitivity literature and aid effectiveness principles highlight the benefits of joint analyses (within governments, between donors, between INGOs, etc.) to generate shared understandings and joined-up working and to improve the coherence of different actors’ programming (e.g. Conflict Sensitivity Consortium, 2012). This also helps overcome the constraints of limited funding and staff (Shahab Ahmed, 2011). The New Deal fragility assessments propose a model of joint conflict analysis that is chaired by the host government and that includes not only the typical development actors but also the private sector and academia (Scott & Midgley, n.d.).

Cross-agency macro conflict analysis

Cross-agency conflict analyses were commissioned in Kenya, Sierra Leone and Sri Lanka by their respective national members of the Conflict Sensitivity Consortium. By pooling intellectual and information resources, it was possible to generate shared understandings. Each consortium member considered the analysis ‘much stronger’ than what they could have developed individually.

Source: Conflict Sensitivity Consortium (2012: 5-6).

OECD (2012: 78) suggests the following questions to help choose a conflict analysis tool:

  • Does the tool’s proposed methodology match the purpose of the analysis?
  • Does the tool’s proposed methodology agree with the ways of working of the evaluation team?
  • Does the evaluation team have the capacity (skills, expertise, access, etc.) to use the tool well?
  • How long does it take to produce a reliable conflict analysis?
  • What are the resource implications of the selected tool (staff time, travel, seminar costs, facilities, data management)?
  • Is the evaluation team able to allocate or secure the required resources?

The purpose of the conflict analysis defines the method. A conflict analysis delinked from strategy / policy / programming processes can be used to build a holistic understanding of the conflict not framed by policy / political priorities. A conflict analysis embedded in a strategy design process (‘strategic conflict analysis’) directly shapes decision-making on what to work on in a conflict-affected country. It helps test and clarify the theory of change and integrate conflict sensitivity into overall strategies (Vaux, 2015). Analysis embedded in project design processes will illustrate if and how a project might engage with conflict dynamics and how to manage that. Conflict analyses can be used to develop principles and limits for activities, to design benchmarks for monitoring and evaluation and to define what conflict sensitivity means in the context, and be developed into an analysis of conflict risk and a conflict prevention strategy (ibid.). Participatory conflict analysis can be used to build a common understanding of the conflict between participants (Fisher et al., 2000). The table below presents levels of analysis needed according to the purpose.

Conflict analysis at the levels of country operational plan, sector and project/programme

Level Conflict analysis
Country operational plan ‘A broad understanding of conflict dynamics and the key drivers of conflict is needed. Conflict sensitivity then involves an assessment of how strategic decisions interact with the conflict factors identified.’
Sector ‘An understanding of how the key issues / driving factors of conflict play out in that sector is needed. For example, in the education sector it may emerge that all teachers are drawn from one ethnic group as they are better educated, that history teaches a very one-sided view of the past, or that antagonism to another group permeates language and literacy teaching. These are issues that relate to individual projects, but can be identified at a sector level, and need to be recognized within individual projects or programmes within that sector.’
Project / programme ‘A more nuanced understanding of the conflict at a micro level is needed. The Do No Harm framework, identifying what divides and what connects people in a context, is one valuable tool for such an endeavour. The analysis at a sector level and at a macro level is pertinent to assessing how an intervention can interact with conflict but not adequate to assess a project’s conflict sensitivity. Conflict analysis at the sector level can also be helpful to inform design choices and indicators at the project / programme level, by identifying issues that affect a range of projects in the sector.’

Source: Goldwyn & Chigas (2013: 17).

Focusing the analysis according to the purpose can ensure the analysis is not too broad. Conflict analyses can focus on different levels –geographically (national, regional, local); sectorally, at the programme / project level; and at the problem level. A local conflict analysis should be informed by a wider analysis but will go down to the level of people and their roles (e.g. as dividers or connectors). A national / international conflict analysis should have a local dimension but will not go into such detail. This leads to two basic types – project-focused and programme- or strategy-focused. The concepts may be the same but the conflict analysis system will be different (e.g. comparing Do No Harm with the 2002 DFID tool).

Ongoing conflict analysis: updating and timing the analysis

The literature widely identifies that conflict analysis should be dynamic and occur on an ongoing basis to refine the analysis, to update to changing situations, and to support consistent monitoring (e.g. GPPAC, 2015). The major distinguishing features of ongoing conflict analysis are that it: focuses on the most critical / relevant issues or questions; tends to start with strong foundational knowledge of the conflict, either from seasoned local staff or from international staff who are deeply engaged on the conflict issues; and tends to generate short, regular, often informal outputs / updates, rather than stand-alone reports.

The World Vision MSTC tool (2015: 62, 113) – which focuses on bringing local actors and participatory methodology (workshops) into country-level conflict analysis – is designed to be cyclical, repeated and ongoing. MSTC workshops, which result in a final conflict analysis report, are planned for every three to ten years, depending on contextual changes and organisational needs. In very changeable settings, or following extreme events, this period can be shortened. The guidance note plans that, in between the workshops, the MSTC report will be updated through context monitoring, focusing particularly on the elements that are most likely to change – such as trigger events and scenarios. It recommends setting up a context monitoring team with ‘MSTC-trained people’ to collect and analyse data (ibid.).

Although not focused on conflict analysis per se, two recent initiatives in the development community – Thinking and Working Politically (TWP) and Doing Development Differently (DDD) – highlight the importance of ongoing political analysis. Hudson, Marquette & Waldock (2016: 1) outline two ‘steps’ and leading question for everyday political analysis:

1. Understanding interests: What makes people tick?

  • Is what they want clear?
  • Are they acting in line with their core beliefs?
  • Do you understand the constraints they face?
  • Is it clear who and what the key influences on them are?
  • Is their behaviour being shaped by social norms about what is appropriate?

2. Understanding change: What space and capacity do people have to effect change?’

  • Are they the key decision-maker?
  • Do they have potential coalition partners?
  • Are their key decision points clear?
  • Is their framing of the issue likely to be successful?
  • Are they trying to achieve multiple things at once?

However, there are many practical challenges to carrying out ongoing analysis. For example, significant time and resources are required. Goldwyn and Chigas (2013: 17) propose a ‘good enough conflict analysis’ can be based on quicker methods – such as workshops including multiple stakeholders, rapid interview processes or a desk study drawing on analyses by other agencies. Another common challenge is that there may not be capacity / knowledge to update the paper in-house, especially if external consultants carried out the report, there was not a transfer of knowledge during that process and if the consultant is not available to update the analysis (Shahab Ahmed, 2011).

Timing the analysis. When conflict breaks out or escalates suddenly, a rapid conflict analysis may help in getting a sense of what is happening. In other cases, an analysis might be commissioned a year or more before an election that could become a conflict trigger. While the literature highlights that strategic and project analyses should be planned according to the policy timeline, Shahab Ahmed (2011) found PCIAs in Pakistan were often commissioned outside of the project cycle and that it was hard to integrate them. Evaluations of DFID’s Strategic Conflict Analysis tool found that ‘poorly timed analyses have failed to be integrated into country planning processes, thus limiting their influence’ and that incentives within DFID ‘militated against initiation’ (Barakat & Waldman, 2013).

Methods and data

Conflict analyses typically combine different methods, including literature reviews and secondary data analysis; participatory methods – community consultations, workshops with project staff and experts; data collection – surveys, media monitoring; and key informant interviews.

Primary research, particularly participatory approaches at local level, should be budgeted for and planned in advance to ensure there is time and funding. Primary research in conflict-affected countries can be extremely costly, be dangerous to carry out and take a long time to produce results, owing to access and safety challenges, the dynamic changing environment, lack of personnel with language skills and local knowledge, etc. Again, potential biases should be identified. A key challenge lies in examining and representing the views of the ‘opposition’, which may in some cases be underground. This can be done to an extent by involving or interviewing proxies.

Desk-based research can provide a broad understanding of the issues, how they interrelate, a historical framing and linkages with academic and policy debates. A range of sources can be found in published literature and grey literature; media reports can be used to track recent events. Two important limitations are that the majority of literature on conflict and developing countries tends to be published by external actors – often donors, think tanks and NGOs (e.g. human rights organisations). This potential for bias should be highlighted in the review. Also, desk-based research lacks an up-to-date perspective and typically has a weak understanding of the local realities of the conflict.

Concerns over sensitivity, security and political implications mean many conflict analyses are not published, particularly those based on primary research. This greatly limits the development of shared understandings and the ability to evaluate the impact of the analyses.

There are a number of useful secondary data sources online. Data from conflict-affected contexts are typically limited and illustrative rather than comprehensive. Triangulation across different sources can improve reliability. Data should be disaggregated to reveal more nuanced information about specific groups – for example sex, ethnicity, age. Gleditsch, Wallensteen, Eriksson, Sollenberg & Strand (2002) explain that the problem with relying on events and statistics to define conflict is that conflicts are experienced in different ways by different people, and conflict events – large or small – can have variable local and international implications. For example, a key global dataset on armed conflict – the Correlates of War project – did not include the Northern Ireland conflict in its dataset as the number of battle deaths per year did not reach the 1,000 deaths threshold set up the project, despite the situation being experienced both at home and abroad as a conflict (ibid.: 17).

Illustrative examples of data sources on conflict

Example of a global monitoring system: Armed Conflict Location and Event Data Project (ACLED)

ACLED tracks political violence and protest data for developing states. Including: dates and locations of political violence and protest, types of event, groups involved, fatalities and changes in territorial control. Also data related to: battles, killings, riots and recruitment activities of rebels, governments, militias, armed groups, protesters and civilians.

Example of a subnational monitoring system: Bangsamoro Conflict Monitoring System (BCMS)

The BCMS is a subnational conflict monitoring system that tracks the incidence, causes and human costs of violent conflict and violent crime in the proposed areas of the Bangsamoro (currently part of the Philippines). It sources data from government and civil society to improve credibility and robustness, principally Philippine National Police reports, supplemented by media reports. Multi-Stakeholder Validation Groups also generate and validate data. The BCMS is a partnership between International Alert, the World Bank and three local academic institutions. It has been combined with the Southern and Eastern Mindanao Conflict Database to form Conflict Alert.

Triangulation and the nuancing and balancing of findings are crucial. Actors have different experiences, perspectives and histories of the conflict and will remember events with different meanings and emotions (Fisher et al., 2000). ‘Facts’ and perspectives are highly politicised. To mitigate this, the conflict analysis process should engage a range of different stakeholders with different perspectives (Conflict Sensitivity Consortium, 2012). External actors should be cautious of what information they select and how they use/reproduce it, as it can have legitimising (and de-legitimising) impacts on certain actors or issues.

Inclusion of gender perspectives in conflict analysis enables a more nuanced and effective understanding of conflict factors, actors and dynamics. It can highlight the gendered nature of the causes and impact of conflict, providing a deeper understanding of the structural issues that need to be addressed through peacebuilding (Anderlini, 2006). Gender variables are, however, often missing from conflict analysis and conflict assessment frameworks. Many conflict analysis frameworks mention gender issues – for example the need for women’s participation in consultative processes or for understanding of the role of gender in social exclusion – in only a cursory sense (ibid.).

Resistance to undertaking gender-sensitive conflict analyses is partly fuelled by the lack of rigorous evidence that gendered approaches make a significant difference to the quality of interventions in conflict-affected countries (ibid.). Guidance on carrying out gender-sensitive conflict analysis is not well developed. The term ‘gender’ is still often used synonymously for ‘women’, resulting in the failure of gender analysis to acknowledge that gender is relational and that men also possess gender identities (Sudhakar & Kuehnast, 2011). Acknowledging men as gendered subjects makes it possible to ask men and women similar questions in gender analysis, and to understand what conflict and peace mean to different women and men (Myrttinen et al., 2014). A gender-relational approach to gender and conflict analysis should include how gender difference intersects with other identities in shaping and being shaped by violent conflict – and in providing opportunities for transformative change (Myrttinen et al., 2014; El Bushra & Sahl, 2005).


‘Evaluation teams are primarily concerned with conflict analysis from two perspectives. First, in assessing “relevance” it will be important to understand whether and how a programme implementation or policy development group developed their understanding of the conflict and context. In other words, what was the basis for their determination of priorities at the policy level or programme directions. Second, in order to assess the “impacts” of policies or programmes, the evaluation team needs to understand the conflict that programmes and policies are attempting to influence or change. An evaluation team thus needs to understand the different approaches to, and tools for, conflict analysis to be able to review the adequacy of the analysis performed or conduct its own analysis if one does not exist’ (OECD, 2008: 68).

Examples of published conflict analyses