Instability in the Kenya–Somalia–Ethiopia Borderlands

Borderlands have been historically dismissed by those in national capitals as peripheral and treated as buffer zones where insecurity could thrive. However, over the last decade, there has been growing attention within peacebuilding and development policy on the ‘central’ role played by borderlands in national security, ‘conflict, statebuilding and development’. It could be argued that the strength of a country’s identity and cohesion is best measured by understanding the nature of the state’s relationship with its borderlands. These areas can be zones of economic opportunity, but they are also places where the distinction between legal and illegal can be blurred and where transnational conflict economies can thrive.

In the Horn of Africa (HoA), borders drawn by former colonial powers rarely align with local imaginations of identity. Cultural and linguistic ties across international boundaries are often stronger than between those living within the same nation-state.1.  While these borderlands should not be considered marginal or ungoverned, they do face unique challenges.

Deris Wanaag (meaning ‘good neighbourliness’ in Somali) was a three-year UK- and Dutch-funded project implemented in the Kenya–Somalia–Ethiopia borderlands between 2023 and 2026. The initiative sought to enhance peace and stability in these areas through integrating humanitarian, development and security interventions.

Avoiding state-centrism, Deris Wanaag worked across the Kenya–Somalia–Ethiopia borders, adopting a systems approach in order to understand and analyse violence. Violent extremism was understood as one component of a broader conflict system, including overlapping threats from public disorder, communal violence, crime, smuggling and other forms of insecurity. The project responded by tackling the system as a whole, collectively addressing the structural drivers, social norms and individual incentives supporting violence.

As the project concluded, a one-day research conference was held at the Rift Valley Institute headquarters in Nairobi on 12 February 2026. Convening practitioners, policymakers and researchers in equal numbers, the event was designed to build on project findings through debate, to maintain focus on the borderlands, and to promote uptake of research findings by a wider audience. It also provided an opportunity to consider the shifting role of research and evidence in peacebuilding and security sector programming.

The timing of the conference emphasised the challenges around the Kenya–Somalia–Ethiopia borderlands. In February 2026, drought conditions were returning to the borderlands following failed rains in late 2025. Concerns were growing in Somalia and Ethiopia over upcoming elections, as well as political tensions in Kenya. In Somalia, relations between the Federal Government of Somalia (FGS) and Somalia’s Federal Member States (FMSs) were arguably at an all-time low. Geopolitical tensions were also escalating following Israel’s recognition of Somaliland. And in the weeks leading up to the conference, there was increased violence on the Kenya side of the border by the Al-Shabaab Islamist militant group. More attacks (including IEDs, vehicle hijackings and small arms attacks) were recorded in January 2026 than in any single month in more than two and a half years.2.

The conference was structured around four panel discussions, each followed by audience interaction. The discussions comprised:

• The political economy of the borderlands.
• The links between climate, migration, livelihoods and conflict in the borderlands.
• Gender and conflict in the borderlands.
• Achieving impact through research.

This Insights Paper is built on these four themes, and seeks to identify the key debates, challenges and research gaps that emerged. Where possible, practical recommendations for future programmes are identified within the context of rapidly evolving international engagement. Unless otherwise indicated, statements in this paper reflect points raised during the discussions. None of the statements are attributable to any specific individual or organisation.