In the ‘golden age’ of mediation, immediately after the end of the Cold War, peace agreements aimed to provide definitive answers to the questions which led societies down the road to violence and civil war. ‘Comprehensive peace agreements’ sought not only to stop the immediate fighting but to revise the fundamental nature of the state and society to make the resurgence of violence impossible.
The Dayton Agreement, for example, which ended the war in Bosnia-Herzegovina, contained provisions related to the protection of human rights, treatment of displaced persons, preservation of national monuments and restitution of property. It also included the Constitution of the State of Bosnia-Herzegovina as an annex.
Such detailed agreements are correlated with long-term reductions in violence (documented, for example, by Joshi and Quinn). However, the heyday of detailed comprehensive agreements is long over. The opportunities for negotiating comprehensive peace agreements are vanishingly rare today due to the nature of societies in which conflicts are taking place and a lack of consensus in the international system. Most intra-state conflicts today involve multiple conflict parties, sometimes with weak command-and-control structures, fighting in fragile states. Many conflicts also have a ‘proxy war’ dimension or are otherwise impacted by the involvement of external actors.
The difficulty of reaching comprehensive agreements – and their limitations – has in some instances led to the negotiation of ‘framework agreements’, which outline the details of a process through which unresolved issues would be addressed without necessarily agreeing upon content. Such agreements perform an important function in contemporary conflict resolution, but, as we will see, are rarely smoothly implemented. More recently, even agreement on frameworks has proved elusive, leading to ‘partial’ agreements to halt or reduce violence without addressing the root causes of the conflict. Such agreements often take the form of local ceasefires or temporary truces, as seen in Syria and Yemen. Although not necessarily damaging in and of themselves, these may have negative side-effects (such as displacement) and reduce interest in an overall settlement to the conflict.
It is essential to find a way to ensure that decades of death and destruction do not give way to an unending succession of corrupt, self-serving governments that prey upon and suppress their populations. This requires mediators and conflict parties to chart a course between the quixotic search for a perfect ‘comprehensive’ agreement and the negotiation of often utopic ‘framework agreements’, or unsatisfying ‘partial’ agreements.
Doing so requires a reconceptualisation of mediation as an iterative process, which seeks to produce a series of micro-agreements (formal or informal) over time and must engage an evolving (and hopefully increasingly inclusive) set of constituencies. In today’s complex environment, this long-term, adaptive perspective offers a viable path to a sustainable and equitable end to violence. Applying such an approach requires significant innovation, not only in the way agreements are conceived, crafted, implemented, and evaluated, but also in the way mediation teams and donor support are structured.
The benefits and limits of ‘framework’ agreements
Despite their potential benefits, comprehensive agreements have important limitations. Peace processes are all necessarily exclusionary to some extent: even in cases where more inclusive dialogues are attempted, powerful veto players are given the principal say on key provisions of the texts. This is normatively problematic, especially if the agreement establishes long-term principles on the structure of the state.
Moreover, codifying long-term governance principles in the heat of conflict encourages the creation of systems which accentuate inter-group differences and make cross-party coalitions more difficult. Both the Dayton agreement in Bosnia-Herzegovina and the Ta’if agreement in Lebanon, for example, created fixed confessional or ethnic systems for dividing power that cemented cleavages, which remain in place today. Given these factors, one should not lament the demise of comprehensive peace agreements too emphatically.
However, this is not to recommend that peace agreements only address the steps necessary to halt immediate violence. The signature of a peace accord can provide a once-in-a-generation opportunity to transform society. This opportunity should not be wasted by simply agreeing to pause fighting and divide up government posts between the warring factions.
One solution to bridge the gap between comprehensive and partial agreements has been the negotiation of ‘framework’ agreements that put an immediate stop to the violence and outline the process through which the broader post-conflict political settlement will be established. An early example of this approach is the 1991 Framework for a Comprehensive Political Settlement of the Cambodia Conflict, otherwise known as the Paris Peace Agreement, which created a mixed international-domestic transitional administration, and outlined the process of getting to elections and a post-conflict constitution.
Most framework agreements envisage three phases. First, the conclusion of a ceasefire, where parties lay down their arms; second, the establishment of a political transition in which belligerents jointly manage the state; and third, a process of structural reform and statebuilding, typically including constitutional reform and elections, as well as processes of transitional justice, security sector reform, and post-conflict reconstruction and development.
These three phases are interlinked, often overlapping, and may compromise and complicate one another. Smooth and linear progress from one phase to the next is rare. Bell and Zulueta-Fülscher, for example, note that the linear trajectories outlined in the peace agreements reached in Somalia in 2000 and 2002, in Yemen in 2011, and in Libya in 2015 (and, indeed, 2020–21) were not followed. Processes move backwards and forwards between the three ideal conceptual phases of ceasefire, transition, and reform.
The continuous negotiation and renegotiation witnessed in most post-conflict contexts has led some to contend that formal ‘framework’ agreements provide an illusory image of the structured process to be followed. Instead, Bell and Pospisil suggest many post-conflict states exist in a state of ‘formalised political unsettlement’. Recognising that this is the most likely outcome of a negotiation process requires international actors to think creatively about the potential benefits of uncertainty and flux.
Accepting that the terms of the political settlement will be renegotiated and revised reduces the ‘finality’ of any agreement, preventing the emergence of perceived ‘winners’ and ‘losers’. This flexibility allows the accommodation of new groups (including the management of spoilers), adaptation to any shift in alliances, and changes to the priorities and balance of power between constituencies. Such flexibility is particularly useful in contemporary contexts of fragmentation. It also provides opportunities for excluded groups to peacefully press their claims over time, which may lead to gradually more inclusive institutions and systems of governance. Moreover, the assumption that negotiation will remain at the heart of institutional decision-making incorporates a long-term vision of compromise-based politics which may be more appropriate for conflict-prone societies than ‘winner-takes-all’ majoritarian, or even other forms of, democratic systems. However, such an iterative, open-ended approach also requires serious thinking about guarantees and vetoes to forestall the risk of conflict continuing or agreements being made in bad faith, as discussed below.
Mediation in contexts of formalised political unsettlement
Recognising that peace processes are often iterative and require constant renegotiation means that the role of mediators cannot end with the signature of an accord. Indeed, in many contexts, transitions are catalysed without a formal agreement being reached. Such ad hoc transitions do not obviate the need for mediation, negotiation and international support but rather may require more of them. The 2015 Nobel Peace Prize, for example, was awarded to four Tunisian civil society organisations for their mediation of a series of agreements between political parties which allowed the Tunisian transition to move ahead. It entailed neither a comprehensive agreement nor a framework agreement, but a continuous process of negotiation between key constituencies. Mediation can perform an important function even in contexts where there is no formal peace agreement, including cases of decisive military victory or unconstitutional regime change, such as coups and mass social protests.
Mediators also play important roles in distinct phases of a peace and transition process – from ceasefires to political reforms. For example, in Burundi, an initial ceasefire in August 2000 did not include several key parties to the conflict. Through continued efforts they were later convinced to lay down their arms and join the ceasefire. In other contexts, such as the Democratic Republic of Congo following the 2002 Sun City Agreement, mediators have been less effective in engaging those opposed to an initial ceasefire. Keeping the door open for groups to join a ceasefire and contribute to a process of political transition and reform (notwithstanding the potential need to maintain military pressure) may incentivise groups to join a ceasefire process at a later stage.
The power-sharing phase of an agreement is also prone to setbacks and difficulties. The structure of the power-sharing government formed in 2001 in Burundi was subject to disputes which were only resolved by continued mediation efforts and several restructurings of the transitional executive and legislature over four until 2005. In Sudan, the coalition of military and civil leaders charged with managing the transition to democracy in 2019 broke down when the military seized power in 2021; a further clash between erstwhile power-sharing partners plunged the country back into violence in April 2023.
During the third, ‘reform’, phase of a transition, mediators, conflict parties and international supporters of the process must tread a fine line between securing sufficient agreement to allow a process to move forward while allowing the reform and state-building processes enough leeway to make genuine changes. This phase provides the greatest opportunity for the inclusive and participatory processes essential for the reform and transformation of the post-conflict state. However, opening up decision making to new social constituencies demands a careful balancing of public participation with elite vetoes. For example, the highly inclusive National Dialogue in Yemen (2012–14) partly failed due to the opposition of powerful groups from the north and south of the country, the Houthi and al-Hirak respectively, which felt that the process did not sufficiently protect their interests. The constitutiondrafting process launched in Libya with the election of a constituent assembly in February 2014 also failed, partly because its deliberations were divorced from the considerations of the main conflict parties.
Concretely, if we consider, for example, a process of constitutional reform, drafters of an agreement should consider outlining processes of participant selection, the mandate given to the bodies established, and the steps of the process to come (with a realistic timeframe). A framework agreement, unlike a comprehensive agreement, would be careful about overdetermining the content of the new constitution or delineating the governance model to be adopted. This type of approach to peace talks allowed the demobilisation of several armed groups in Colombia and their participation in the 1990 Constituent Assembly. The Paris Peace Agreement in Cambodia and the South African peace process managed the balance between the interests of key conflict parties and the need for broader public participation well.
Lessons for international assistance
Sometimes the careful drafting of agreements may alleviate some of the risks of elite capture and entrenchment leading to conflict resurgence. The use of ‘sunset’ and ‘sunrise’ clauses, for example, which outline when temporary provisions will end and what will then replace them respectively, are useful. However, such text-based approaches will only work if they reflect the existing political and military balance of power. These can sometimes be shaped by sustained pressure and support from domestic and international actors, but safeguarding against reversals, missed deadlines and reneging by parties needs vigilance.
Sustained engagement requires a degree of continuity and unity of purpose from components of the international community supporting the process. In Libya, for example, momentum was lost after the successful conclusion of the Libyan Political Dialogue Forum (LPDF) in January 2021, due to a change in the UN leadership and faltering international support. The Government of National Unity, which was intended to last for only one year, was allowed to entrench itself. While international unity may be elusive, continuity in official mediation teams and strategy where possible, and detailed handovers and explicit rationales for changes in approach where necessary, could be more easily encouraged.
International development assistance and financial aid should be conflict sensitive and linked to progress within each phase of the process. In Sudan, for example, lack of international support to the economy, which might have been linked to certain benchmarks on the inclusivity of the transitional administration, helped contribute to its military capture. While the political and potentially contested nature of processes of national dialogue and/or constitutional reform is often recognised by international actors, it is common for supporters of peace processes (and sometimes conflict parties themselves) to consider other reform processes, including security sector reform, transitional justice, and economic and development programming, as technical – rather than highly political – processes. There is a need for dialogue and mediation capacity to support all processes designed to help rebuild post-conflict societies.
The assumption that the revision and renegotiation of prior agreements is essential and does not constitute a failure of the process requires donors to rethink models of assessment, monitoring, and evaluation of mediation initiatives. With this mindset, the conclusion of micro agreements and continued engagement of parties in dialogue and negotiation is a success in and of itself. Mediators must also be attuned to the possibility that their involvement in a process must be calculated in terms of years, rather than days and months.
On the other hand, sustained international involvement in the internal and domestic political affairs of a post-conflict state may raise concerns about the sovereignty and legitimacy of constitutive processes. As such, insider mediators – those who come from the societies in which they are mediating – should be better supported by external actors, while paying careful attention to the maintenance of impartiality and independence. Insider mediators possess significant advantages, notably legitimacy, commitment, and nuanced appreciation of the micro-dynamics of a society. However, they are also prone to intimidation, manipulation and perceived or real biases. Long-term mediation in contexts of formalised political unsettlement is consequently best achieved through genuine partnerships between domestic, regional and international stakeholders.
Beyond comprehensive and partial peace processes: an iterative approach
If the gap between unattainable ‘comprehensive peace agreements’ and unsatisfying ‘partial agreements’ is to be successfully bridged, mediators must help parties develop adaptive frameworks and design processes to answer the complex questions facing society in a sustainable and inclusive manner. This requires balancing the short-term interests of the conflict parties with the longer-term interest of broader society and future generations. A process which diverges too drastically from the perceived interests of key constituencies is likely to be torpedoed by veto players, as we have seen in Sudan. However, a process which simply reproduces elite-dominated and unresponsive governance models which fail a vast majority of the population is also likely to sow the seeds of future discontentment and conflict. Mediation should seek to maintain this delicate balance as the peace and transition process moves forward.
Mediators and donors must recognise that the linear model of ceasefire to power-sharing political transition to permanent constitution and democratic elections rarely works. The (re)negotiation of a country’s political settlement after civil war is most often an iterative process. For donor support, mediation teams and international post-conflict development and state-building programmes to be designed and evaluated accordingly requires significant innovation. Careful and flexible drafting of peace agreements, diligent monitoring of progress, and sustained, conflict-sensitive and tailored international support will help ensure that formalised political unsettlement, where agreements are adapted and revised over time, promotes long-term prosperity and stability and thus the best outcomes for populations.
