Somaliland's peace conferences

Peace accumulated through agreements

Somaliland's peace process was not one conference and was not the automatic return of an unchanged customary order. A Somali-led mapping project identifies about 39 peace and reconciliation meetings between 1990 and 1997. Local ceasefires and bilateral agreements preceded regional meetings; those meetings made wider conferences possible; and national conferences converted some of the authority used to stop fighting into public institutions. In Sanaag alone, a sequence of thirteen meetings followed initial contacts before the Erigavo conference in 1993. The organizational unit was therefore an accumulating agreement, not an all-purpose summit.¹

That conclusion rests heavily on retrospective participatory research. The Academy for Peace and Development and its partner organizations compiled meeting inventories, collected documents and signed agreements, convened working groups, and interviewed people who had participated in or witnessed the events. The larger peace-mapping study conducted more than two hundred interviews across eleven Somali peace processes. Its researchers also warn that memories were filtered through later developments, that the meaning of success varies by party, and that an accord between two groups can disadvantage a third. Those limitations make the archive unusually rich without turning its conference counts or causal interpretations into a representative outcome study.²

The sequence began amid the collapse of coercive state authority. The Siad Barre government's war against the Somali National Movement and civilians in the northwest left destroyed towns, displacement, armed clan forces, and unsettled deaths and property claims. The SNM's military advantage did not give it accepted governing authority over every northern community, and the movement itself did not control all armed groups. Before a government could reliably disarm militias, people had to know who could make a promise, carry it home, and restrain those expected to comply.¹³

Xeer connected a promise to people able to enforce it

Somali xeer supplied locally understood procedures for assigning responsibility, negotiating compensation, regulating access to water and pasture, and agreeing future conduct. A guurti could refer to a council of senior elders assembled for a problem of corresponding scale. Mediators needed to be acceptable to the parties, the venue needed to be treated as neutral, and lineage groups could share responsibility for compensation through diya-paying arrangements. The rules were unwritten and adaptable rather than uniform: terminology and practice varied, Islamic and customary reasoning overlapped, and new agreements answered new circumstances.⁴

This infrastructure made delegation consequential. An elder or mediator was not valuable merely as a respected speaker. The role mattered because the person could consult a constituency, carry an obligation across a conflict line, and help mobilize compensation or restraint. In Sanaag, agreements restored rights to travel, trade, and use pasture across clan boundaries, and elders communicated terms within and between communities. Agreements did not eliminate conflict; they established a known procedure and a network of responsibility for addressing it.⁴

The same architecture bounded participation. Patrilineal representation made male elders the normal voting delegates even though women connected lineages through marriage and natal kin, raised money, carried messages, organized protests, and sustained host logistics. Custom was neither timeless nor automatically inclusive. The peace-mapping synthesis explicitly cautions that customary practice can innovate, enforce conservative authority, conflict with gender equality, or acquire undeserved legitimacy merely by being called traditional.²

Burco made a polity before it made a secure government

Early contacts in 1990 and local meetings at Oog and elsewhere led to a national meeting in Berbera in February 1991 and then the Burco conference in April and May. At Burco, delegations accepted the SNM's political leadership for a two-year transition and declared withdrawal from the 1960 union with Somalia. The conference joined reconciliation to a claim of public authority, but it did not give the new administration a monopoly of force, a reliable revenue base, or settled rules for succession. Demobilization disputes and competition over Berbera port and Burco soon produced new fighting.⁵

The Sheekh conference in October and November 1992 illustrates how one meeting prepared another. It consolidated a ceasefire, required militias to leave battlefields and roads, called for the return of fixed property and prisoners, and brought clan guurtis together as a national council. The resulting council could assume functions that the SNM's divided central institutions could no longer perform and could prepare a broader conference before the transition expired.¹

Michael Walls interprets the 1991–93 process as a combination of dialogue, accepted mediation, consensus-seeking, and a willingness to defer some of the most contentious compensation disputes. His account is important because it rejects the idea that Borama itself ended the conflict: the interim settlement made later peace and elections possible, but violence resumed. It is also a focused interpretation by a scholar who worked with the Academy for Peace and Development, not proof that every participant understood consensus or deferred justice in the same way.³

Borama joined reconciliation to civilian succession

From January to May 1993, the Borama conference first negotiated a peace and security framework and then a national governing structure. One hundred and fifty official voting delegates were apportioned among clans, while a much larger body of observers, organizers, guests, and hosts surrounded the formal sessions. Some minority clans were not represented. Seventeen women attended as official observers and pressed for a larger role, but the voting body remained male. Delegates adopted a Peace Charter and a National Charter, retained an executive presidency, formed two legislative houses, and selected Mohamed Ibrahim Egal to replace Abdirahman Ahmed Ali Tuur.⁶⁷

The surviving Somali-language National Charter is primary evidence of the formal artifact produced at Borama. The Academy's historical reconstruction and independent legal-political analysis are needed to interpret how that text worked, because an archived charter cannot show the bargaining outside plenary sessions, the degree of acceptance in each community, or how faithfully later officeholders followed it.⁸⁶

Borama linked horizontal peace to vertical succession. A military movement yielded executive leadership through a conference vote rather than another armed contest, and elders' temporary mediating authority became an upper legislative house. The transfer was a substantial institutional achievement, but the embedding of the Guurti also changed a flexible convening role into an office whose composition and renewal could become less responsive than the relationships that had authorized mediators during war.⁶³

Local finance created accountability and leverage

Most conference resources came from host communities, Somaliland businesses, and diaspora networks. Hosts supplied food, shelter, meeting space, security, and time; women were central to fundraising and logistics. Local dependence made participants conscious that delay and failure imposed costs on people they knew, while the absence of a fixed donor timetable left room for consultation, mediation, and revised proposals. At Borama, consensus usually displaced a simple majority vote even though reaching it took months.⁹⁶

Locally financed should not be read as literally free of outside assistance. Sarah Phillips reports roughly $100,000 from two Western donors for Borama. The Academy's reconstruction records United Nations air transport, some bilateral support in kind, and recollections of Norwegian funds. The sources disagree on the exact external total while agreeing that Somaliland and diaspora contributions bore the overwhelming burden. That distinction matters: the credible mechanism is local financial dependence, not the romance of complete isolation.⁹⁶

Finance also concentrated power. Phillips reports that after Borama, Egal borrowed about $6 million from eight Isaaq traders to fund demobilization, build an army, and secure Berbera port. Tax exemptions and other advantages rewarded creditors and helped a small political-commercial coalition dominate economic access. Mutual dependence between political leaders and traders supported order while narrowing opportunity and strengthening executive patronage. Peace and exclusion were not separate phases of the settlement; some exclusion helped finance the coercive capacity that sustained it.⁹

Women made peace work without equal decision rights

Women organized campaigns, prayer meetings, poems, marches, and demands across clan lines. At Sheekh, organizers read a written platform, installed sound equipment when kept outside, promised material support for a police force, and pressed delegates to remain until they addressed the agenda. The seventeen women invited to Borama could speak, circulate pamphlets and poems, and lobby voting delegates, but observer status gave them no vote. A later comparative case study finds that strong women's organizations and economic roles opened the door to observation while male-only elder selection and gender norms constrained influence.⁷

Siham Rayale's narrative research adds a different kind of evidence. Interviews and ethnographic work conducted in Hargeisa, Burco, Borama, and Gabiley recover women's informal organizing and their accounts of violence, sacrifice, and political marginalization. Those narratives challenge any equation of peace with the absence of large-scale war: gendered insecurity and subordination can continue under a stable public order. The purposive interview sample was built to recover missing voices, not to estimate the prevalence of any view among all Somaliland women.¹⁰

Women could also support mobilization for war, and neither women nor men formed unitary political groups. Recognizing women's peace work therefore requires more than adding a uniformly peaceful category to an elder-centered story. It requires tracing who raised resources, crossed boundaries, pressured armed kin, spoke inside or outside a hall, and then received or lacked decision rights in the institutions that followed.¹⁰¹

Hargeisa closed one war under more centralized conditions

Borama did not produce an enduring end to violence. Fighting in Hargeisa and Burco from 1994 to 1996 exposed the Guurti's inability to bind every militia and the new government's contested authority. A diaspora-organized Council for Peace and Development pursued visits, correspondence, ceasefire proposals, and dialogue from April 1995 through January 1997. Its contemporary report documents persistence and failed approaches as well as contribution, but it is a participant account of its own mission rather than an independent evaluation of which intervention caused peace.¹¹

Local reconciliation at Duruqsi and Beer helped bring reluctant parties toward the Hargeisa conference of October 1996 to February 1997. Unlike earlier major meetings, the government funded Hargeisa. Its 315 delegates included sitting legislators plus additional clan representatives, minority representation, and female observers. Participants reinstated Egal, selected a vice president, appointed legislative members, and replaced the National Charter with a provisional constitution. Contemporary complaints that government resources distorted the conference show the changed political economy: an emerging state could now convene and finance peace, but could also influence the forum that authorized it.¹²¹¹

The cumulative process therefore changed its own conditions. Early meetings worked through dispersed authority because no actor could impose a settlement. Later meetings operated alongside growing executive revenue, security forces, credit relationships, and patronage. Phillips argues that this settlement contained large-scale violence partly through locally negotiated rules and elite mutual dependence, but also excluded many who rejected Somaliland's claim of independence, especially in contested peripheral areas. That is an analytical warning against treating institutional endurance as universal consent.⁹

Organizational profile

Authority was simultaneously local and federated, customary, religious, military, commercial, and eventually executive. Decisions moved between community consultation, mediator committees, conference plenaries, elders' councils, and a presidency. Host and diaspora contributions resembled a temporary coalition rather than organizational ownership. Coordination depended on mutual adjustment, ritual and rule, planning committees, and ad hoc teams. Knowledge traveled upward from witnesses and constituencies, laterally through kin and mediators, and back down through delegates expected to explain and enforce agreements.⁴¹

Measurement was comparatively weak: ceasefires held or broke, roads opened or closed, prisoners and property returned or did not, and armed groups complied or defected. Learning came through repeated meetings, revised agreements, mediator experience, and institutional change rather than standardized indicators. Adaptation occurred locally and slowly until crises demanded another conference. Communities, members, public institutions, and later generations benefited unevenly. Capture, suppressed voice, dependence on credible individuals, and renewed violence remained central risks.²

Idea fingerprint

• Purpose, mission, and institutional legitimacy scores 3 because ending reciprocal violence and authorizing a polity were inseparable missions whose acceptance had to be rebuilt meeting by meeting.
• Authority, legitimacy, and acceptance scores 3 because military command, elder standing, clan obligation, commercial leverage, and public office overlapped without one initially dominating the rest.
• Delegation, decentralization, and responsibility scores 2 because delegates carried obligations between conferences and constituencies, although representation was uneven.
• Coordination, communication, and common understanding scores 3 because cross-line messages, consultations, mediation, and nested agreements were the operating mechanism of peace.
• Structure, hierarchy, and scale scores 2 because local, regional, and national meetings accumulated into legislatures and an executive without following one predesigned hierarchy.
• Decision-making, judgment, and bounded rationality scores 3 because acceptable mediators, deferred disputes, consensus, and time allowed decisions under incomplete information and threat.
• Measurement, accounting, and control scores 0 because behavioral signs of compliance mattered, but formal measurement was not a sufficiently evidenced organizing mechanism.
• Cooperation, incentives, and organizational equilibrium scores 3 because reciprocity, compensation, host burdens, creditor bargains, and mutual dependence aligned actors who retained coercive options.
• Work design, productivity, and automation scores 0 because distributed peace labor was vital but productivity or automation was not an organizing concern.
• Knowledge, expertise, and professional autonomy scores 2 because local history, customary procedure, religious reasoning, and mediator credibility supplied expertise outside a professional bureaucracy.
• Learning, quality, and reliability scores 2 because settlements were tested, revised, nested, and sometimes abandoned after renewed fighting.
• Strategy, competition, and adaptation scores 2 because movements, clans, traders, and officeholders adapted alliances and conference rules while competing over force, revenue, and representation.
• Innovation, entrepreneurship, and renewal scores 1 because the hybrid settlement was institutionally inventive, but venture formation or product renewal was not central.
• Governance, stewardship, and accountability scores 2 because charters, legislatures, security arrangements, and public succession arose alongside unresolved accountability and capture risks.
• Culture, informal organization, trust, and voice scores 3 because informal standing and trust made agreements enforceable while gendered and lineage-based rules limited who could speak and vote.
• Executive attention, information, and organizational sensing scores 2 because the SNM and later presidency depended on conference information, mediator reports, and shifting acceptance across armed groups.
• Organizational ignorance scores 0 because missing voices and disputed histories are serious gaps, but the evidence does not isolate ignorance production as a distinct mechanism.

Comparisons distinguish three forms of public deliberation

Botswana's kgotla assemblies operate as continuing public forums within an established state; Somaliland's conferences first had to produce enough security and authority for ordinary public institutions to function. Tunisia's National Dialogue Quartet mediated a constitutional transition through organized civil society within an existing state, while Somaliland's mediators worked among armed lineage constituencies after state collapse. Namibia's communal conservancies federate local resource governance through statutory recognition; Somaliland's conference sequence used local agreements to help create the statutory order itself.

Research gaps

The most important remaining work is controlled by people whom published English-language accounts reach unevenly: Somali-language agreements and oral histories preserved by host communities; accounts from women who did not become recognized activists; minority, occupational-group, youth, displaced-family, and eastern-community perspectives; and testimony from people who opposed Somaliland independence. Financial ledgers could test claims about conference funding, host burdens, compensation, disarmament, and later creditor privileges. Longitudinal research could distinguish which agreements remained enforceable, which relationships failed, and when a flexible mediator role became an unaccountable office.