A Multi-Disciplinary Approach to Reducing GBV in Somalia

Gender-based violence (GBV) is rooted in cultural norms, power imbalances, and the social performance of gender. These factors shape how violence is used to exert control, enforce gender roles, and perpetuate inequality. In conflict settings, GBV can be used as collective punishment and surrogate violence, intended to fracture communities or exert control over rival groups. Conflict-related sexual violence (CRSV) emphasizes the structural and communal dimensions of violence specifically related to conflict, beyond individual incidents of sexual and gender-based violence. Thus, any effort to effectuate a sustainable and durable reduction in CRSV must not only respond to individual instances of assault but address the risk in its full complexity.

Results-based protection (RBP) offers a roadmap for an interconnected, system-wide approach to reducing protection risks. A case study from Abudwaq, Somalia, provides a clear example of how multidisciplinary strategies and community-driven solutions can reduce the risk of CRSV without relying solely on traditional GBV programming, which focuses on immediate intervention that addresses the physical, health, and psychosocial needs of survivors. Traditional GBV programming may also seek to prevent GBV by focusing on the root causes of gender inequality and systemic discrimination or on reducing the potential for exposure, often with the creation of women and girls-friendly spaces, site safety audits, or behavior change communication regarding gender norms. These efforts, although critical, are long-term and will often not prevent or stop CRSV during ongoing humanitarian crises.

Despite widespread endorsement of the Centrality of Protection, the humanitarian sector has developed few concrete approaches to translate it into effective action. Results-Based Protection (RBP), innovated by InterAction and endorsed across the humanitarian community, is one of the only approaches that offers a practical pathway to reducing protection risks through outcome-oriented analysis and multidisciplinary strategies. Rather than centering activities or service delivery, RBP emphasizes outcomes—what meaningful change looks like for affected people—and works backward to determine how to get there. The RBP framework is explained in three parts: context-specific protection analysis, multidisciplinary strategy building, and the use of outcome-oriented approaches. It posits that any meaningful reduction in a given protection risk is predicated on a detailed understanding of the risk patterns in a specific context as experienced by the affected community. Building this understanding necessitates engagement with the affected population to identify the components of risk: understanding the specific threats (e.g., the perpetrator), who are vulnerable to these threats and why, and the capacities of affected people to cope with and reduce the threat and their own vulnerabilities to the threat.

In Abudwaq, with guidance from InterAction, partners including the World Food Program (WFP) and local NGOs Himilo Relief and Development Association (HIRDA) and Active in Development Aid Somalia (ADASOM) undertook a participatory research project, investigating mutually reinforcing negative cycle of food insecurity, conflict, and protection risk.

The action-based research (ABR) consisted of a two-part methodology that involved participatory data collection and analysis with affected groups, ensuring that knowledge was generated through the direct and active involvement of those with lived experience of the issues and those who seek to address those issues. In this way, the research went from being something that “experts” do about a group of people to something done with people. In the initial research stage, community members joined a series of focus groups disaggregated by gender, age, and displacement status. They engaged in a day-long series of structured conversations and participatory methods that allowed them to explore the linkages between food insecurity, conflict, and protection risk. The specific data collection exercises included Rivers of Life, which generate rich data on individual experience; Problem Trees, which enable group dialogue and analysis of causes and consequences; Community Mapping; and the Risk Matrix, allowing for both dialogue and analysis of protection risks. In the overall project methodology, participatory research was followed by collective analysis and risk prioritization among the project partners. Findings and pathways for risk reduction were again validated with community members.

In Abudwaq, a remote, arid district bordering Ethiopia, the community faces a complex and protracted situation of violence. Over decades, competition for land and resources has led to ongoing clan conflicts, particularly between the Reer Siyaad Xuseen, Wagardhac, Saleeman, Sacad, and Dir clans. This violence is driven by cyclical competition among pastoralists, compounded by the climate crisis and the collapse of livestock populations, which has heightened competition over scarce resources.

Community members identified a novel form of CRSV engendered by local conflict dynamics. Because the clan conflict predominantly targets men for reprisal killing, male household heads have begun displacing to urban areas or clan militia camps, leaving women and adolescent girls, who traditionally herded small livestock like goats, to take on culturally masculine roles in herding larger livestock such as camels and cows. Where male herders have been targeted for cattle raids or assassination, female herders have been increasingly subject to sexual violence by rival militias as a form of retribution for the broader clan conflict, especially when they graze their herds on contested lands or cross fluid clan boundaries in pursuit of lost livestock. A woman who reported being displaced due to the clan conflict described how “these sexual attacks serve to instill fear and control. They are a deliberate attempt to intimidate our community. This tactic is aimed at facilitating forceful land grabs by the militias.”

Community members observed that, while police are trusted resources when
it comes to non-conflict-related instances of sexual assault, militia attacks on female herders are more likely to be reported to clan leaders and militia, as violence associated with clan conflict is perceived as falling outside of state justice mechanisms. At the same time, the kind of traditional systems that adjudicate redress for sexual assault within clans and sub-clans (Xeer) are rarely harmonized between clans, while the social norms that proscribe the targeted killing of women by clan militia also ensure sexual violence is not prioritized among leaders as needing negotiated de-escalation. The intersectional nature of this risk, falling between state and traditional justice systems, results in an expectation among survivors that remedy can only be sought by reciprocal sexual assault on a woman from the offending tribe. Men and women noted this issue of CRSV, with a male respondent observing, “when our women are raped, they urge another female to be raped from the perpetrator’s clan in the way of retaliation and sending a message. It’s like tit for tat.”

A core strength of the RBP framework is its ability to incorporate multidisciplinary strategies to address complex protection. Simply put, in understanding the nature of a protection risk from the community’s perspective, RBP encourages humanitarians to consider what actors are necessary to integrate their responses and effect change. Even as the community defined the risk of CRSV, it also identified pathways for mitigation. Noting that women and girls’ safe spaces and GBV response programming already existed and were not culturally relevant to this unique risk of CRSV, community members instead emphasized the need for multidisciplinary approaches that better situated the risk in the context. They focused on how the risk could be reduced by concentrating on preventing exposure of women and girls to violence through engendering improved access to resources and establishing more effective access to justice by enhancing negotiation between clans. Once aligned with the protection risk equation, their input formed the basis for a more comprehensive and proactive strategy, that would alter all elements of the risk, including: reducing the insistence on reciprocal sexual violence as a source of threat; reducing exposure to violence among female herders by opening alternate grazing sources; and increasing the capacity of communities to use existing peace mechanisms to address sexual violence as well as other forms of conflict-related violence targeted at men.

Programming along each of these pathways necessitates composing a tapestry of actors from across the humanitarian and development sectors, each with specific strengths necessary to contribute to a meaningful reduction in the risk of this specific reciprocal CRSV. The specific change-pathways comprised the following actions:

1. Negotiation and Conflict Resolution – Threat Reduction Multidisciplinary actors—conflict resolution experts, humanitarian protection and GBV specialists, and local leaders—work to promote cross-clan dialogue. Through these negotiations, women will be brought into the conversation to address critical issues like lost livestock and sexual violence. The focus on negotiation will be key to reducing resource competition and establishing common Xeer frameworks, to both reduce a central driver of CRSV and support a common response framework. The role of women in negotiations will be explicitly addressed to ensure survivors perceive themselves as having recourse to a legitimate justice mechanism that does
   not include furthering cycles of CRSV.
2. Environmental Management and Resource Access – Vulnerability Reduction Environmental experts and protection actors will collaborate with communities to secure additional grazing lands, reducing the need for women to cross dangerous clan lines. In this context, HLP-focused protection actors are best positioned to lead, given their experience with land rights and dispute resolution. While development actors typically play a role in land management elsewhere, their presence in Abudwaq is minimal due to the area’s remoteness and operational constraints. At the same time, community-based protection specialists will work with women herders to build information-sharing networks and early warning mechanisms with clan leaders and militia members and ensure that women have access to information pertaining to shifting clan boundaries and threat postures between militias. By using participatory mapping and cross-clan land management, the project will address both the scarcity of resources and the risk of opportunistic violence that women face when herding.
3. Strengthening Community Self-Protection – Capacity Strengthening – Building on existing community self-protection (CSP) strategies, the plan of action supports women in developing their own protection plans, informed by early warning systems and risk analysis. Community-based protection actors will lead this work, drawing on their local networks and contextual understanding to support women in developing protection plans that reflect real threats and coping strategies. Where possible, they will be supported by organizations with expertise in women’s empowerment and risk analysis, though such actors are limited in this remote and highly patriarchal context
   where conflict and displacement have deeply eroded traditional support structures. In parallel, efforts to strengthen community policing will be explored through advocacy with governance and civil-military actors, aiming to expand the role of local police—who are already seen as trustworthy in non-conflict-related cases of sexual violence—to include more consistent and effective responses to CRSV.

These multidisciplinary efforts are the backbone of the RBP approach in Abudwaq, showing that a combination of environmental management, conflict resolution, and community self-protection can produce tangible reductions in the risk of CRSV.

It must be noted that the elements of the protection risk equation are somewhat fluid. Although supporting women to engage in clan negotiations is without question an augmentation of their existing capacity, it has been situated as reducing the threat of calls for reciprocal sexual violence. Similarly, improved policing will certainly impact the behaviors of perpetrators in militias, but it also offers a framing of existing community capacity to respond to risk in a new way.

Of the three elements that define results-based protection—context-specific
protection analysis utilizing the protection risk equation, multidisciplinary
strategies, and outcome-oriented methods—it is the last that is most fundamental to understanding the approach. By engaging with outcome-oriented methods throughout the project cycle, humanitarians can be clear on the protection outcomes they are trying to achieve, have a theory about how to achieve those outcomes, and plot a course of action toward them. Outcome-oriented methods prompt practitioners to continue learning and adapting responses in light of new learning and changes in the context.

While not the entirety of an outcome-oriented approach, monitoring and evaluation are essential components that enable practitioners to track progress toward protection outcomes, surface new learning, and adapt interventions in response to shifting risks. The success of this approach will be evaluated through the lens of the GBV Prevention Evaluation Framework (GBV PEF). The PEF is a bespoke version of RBP, tailored to measuring GBV prevention outcomes. Although it addresses all elements of RBP across the project cycle, the Abudwaq response plans to utilize specifically its monitoring, evaluation, and learning (MEAL) guidelines designed to help non-GBV actors contribute to reducing GBV risks. In Abudwaq, the project will use this framework to assess how changes in resource access, community protection strategies, and cross-clan negotiation reduce the risk of CRSV.

Key metrics will include:

• Changes in the behavior of perpetrators (Threat): Tracking incidents of CRSV in grazing lands and the reduction of violence in areas with enhanced police presence and negotiated agreements, measuring the recourse to negotiated Xeer when instances of CRSV do occur, and engaging women in clan negotiations. It must be noted that RBP does not endorse solely relying upon incident tracking as a source of measurement but suggests that it should be one element of measuring changes in the behavior of threat actors and community perceptions of threat.
• Improved access to resources (Vulnerability): Measuring the number of women with safe access to grazing land without crossing contested clan lines.
• Community engagement in self-protection (Capacity): Evaluating the adoption of community protection plans and their effectiveness in reducing exposure to opportunistic violence.

Methods such as results journals—maintained by women and other community members—and the Most Significant Change technique, which uses storytelling to capture shifts in experience and behavior, offer participatory ways to track progress toward GBV risk reduction. By embedding the GBV PEF into the evaluation process, the project ensures that its outcomes are measurable and grounded in a broader understanding of reducing GBV risks through multidisciplinary action. This approach helps bridge the gap between reactive responses and proactive risk reduction, contributing to long-term protection outcomes. At the same time, given the emphasis of the methods championed by the GBV PEF on engaging communities to monitor behavior change, this approach empowers affected people to be attuned to risk reduction and their role in effecting it. While promising, the feasibility of these methods in highly insecure or remote areas, where literacy levels may be low and social cohesion fractured, will require careful adaptation and sustained accompaniment by trusted local actors.

CRSV, like any protection risk, is a complex phenomenon, impacted by myriad drivers and compounding factors. It is unlikely that any single actor or even a limited number of humanitarian agencies can reduce this risk. Risk reduction requires collective action. Thus, the GBV PEF and RBP measure the contributions a given project makes toward a protection outcome, rather than assuming the project will fully ameliorate the risk. To be effective, however, the project must be seen as a part of a broader interagency strategy aimed at reducing the same protection risk. It is critical that any project clearly understands and articulates its limitations within the protection risk equation-based strategy. In Abudwaq, for example, a potential response area could be targeting survivors of CRSV with specialized case management approaches or working to increase the amount of arable land suitable for grazing, among numerous others. The MEAL plan for this project will seek to observe contributions toward the outcome of reduced instances of reciprocal CRSV and situate our impact in the broader context. The MEAL strategy will consider the actions of other humanitarians, peacebuilders, state and clan actors that could, even unintentionally, also impact risk reduction.

In any multi-agency approach to measuring risk reduction, even within a single sector, harmonizing outcome indicators and measurement approaches across diverse actors poses both practical and methodological challenges. Each agency brings its own MEAL systems, resources, and comfort levels with qualitative methods, and these differences are especially stark between INGOs and NNGOs, and between humanitarian and peacebuilding actors. Many of the outcome-oriented methods proposed represent a significant shift for humanitarian MEAL teams accustomed to quantitative, activity-based tracking. This was evident even during the joint protection risk analysis, where inconsistencies in facilitation style, note taking, and willingness to engage on sensitive topics impacted the depth and comparability of data. InterAction led the initial analysis on behalf of all partners. However, sustaining that level of coherence throughout implementation will require deliberate decisions about who collects and analyzes qualitative data, how to ensure quality and comparability, and whether analysis is centralized or agency-specific. Without such coordination, the power of outcome-oriented, community-driven measurement could be diluted.