The General Framework Agreement for Peace in Bosnia and Herzegovina was initialled at the Wright-Patterson Air Force Base in Dayton, Ohio, in November 1995 and signed in Paris that December. Its thirtieth anniversary arrives with Bosnia in its deepest political crisis since the war it ended. Milorad Dodik, the long-serving leader of Republika Srpska, was convicted in early 2025 for defying the international overseer, stripped of his office that summer, and a loyalist won the snap election to replace him that November. The High Representative who pursued him, Christian Schmidt, has since announced his own departure, his legitimacy never recognised by Russia or China. The anniversary therefore finds Dayton in a characteristic posture: durable enough that no one expects a return to shelling, exhausted enough that almost no one defends it as a model of government.
This is precisely why the agreement is worth revisiting now, and not only as Balkan history. Dayton remains the most frequently cited template for ending an intractable ethno-national war, and it is invoked, loosely and often, in discussions of Israel-Palestine. The temptation is understandable. Here was a conflict among communities claiming the same land, settled not by the victory of one side but by an externally brokered architecture of separation, power-sharing and international supervision. The question is whether citation amounts to applicability. What follows examines four of its core mechanisms, the entity structure, the Office of the High Representative, ethnically defined representation, and the constitutional court, moving for each from what it did across thirty years in Bosnia to what it would, or would not, do in Israel-Palestine. The examination yields warnings, but warnings are not a program, and since the status quo in the Holy Land is itself the most dangerous option available, this article does not stop at caution. It closes by proposing what the Bosnian experience suggests a serious framework would have to do differently.
Three disanalogies govern everything that follows, and honesty about them is the price of invoking Bosnia at all. The first is asymmetry. Dayton ended a war among armed combatants and, though they were far from equally responsible for it, treated them as co-equal parties to a ceasefire. Israel-Palestine is not such a war but an occupation, one side sovereign and armed, the other stateless and living under it, so that every imported mechanism lands on radically uneven ground. The second is contiguity: Bosnia was one territory whose front line could be frozen into a border, while the Palestinian territories are two non-contiguous fragments, Gaza and a West Bank broken internally by settlements. The third is permanence: arrangements sold as interim tend to harden into the thing they were meant to precede, a pattern Bosnia illustrates and the Oslo process has already lived. Each of the following sections reviews these three elements.
The Entity Structure
Dayton’s foundational act was territorial. It divided the country roughly 51 to 49 between a Bosniak-Croat Federation and Republika Srpska (Serb Republic), drew an inter-entity boundary line along the war’s final front, and left a deliberately weak central state above the two. The arrangement worked, in the narrow sense that it stopped the killing. But it did so by ratifying a map that three years of expulsion had already produced. Dayton built its success on the fact that large territories had been ethnically cleansed; without that prior separation, two entities each with a clear majority could not have been drawn at all. American diplomat Richard Holbrooke, who imposed the structure, never disowned this charge. He regarded the splitting of the country into two entities, and the decision to let the Serb entity keep the name given to it by the Bosnian Serb leader Radovan Karadžić, later convicted of genocide, war crimes and crimes against humanity at the ICTY, as the agreement’s core weakness.
The analogy to Israel-Palestine is the first element approached and the first that fails. Both are cases of two peoples and one contested territory, and separation has long been the dominant grammar of proposed solutions.
Partition is not an idea Bosnia exports to Israel-Palestine; it is one this land has entertained and set aside before. The Peel Commission proposed it in 1937 and the United Nations voted it in Resolution 181 in 1947, and each attempt ran aground on the same fact Dayton would later confront, that the populations a line was meant to separate were already too entangled to sort cleanly. The federation and confederation frameworks now under discussion, Whitbeck’s among them, are responses to that recurring failure, and they concern the whole territory between the river and the sea, not the West Bank alone.
Dayton’s entity map presupposed separability, and that is exactly what this land lacks. The populations are interpenetrated rather than sorted: settlements threaded through Palestinian space, Jerusalem claimed whole by both, mixed cities inside Israel proper. A Dayton-style entity line could be drawn only by accepting the settlement enterprise as a finished fact, which reproduces the very objection levelled at Dayton, that it rewards what was taken by force, or by moving people on a scale no one will openly propose. The clean line between Republika Srpska and the Federation has no honest equivalent across the Green Line. Worse, there is no single map to redraw. The Inter-Entity Boundary Line traced one continuous front, whereas the Palestinian territories are two severed fragments, Gaza cut off from a West Bank that is itself broken into enclaves by settlements, bypass roads, and the closure regime. An entity line presupposes not just separability but contiguity, and the geography of the case offers neither. This is where John Whitbeck’s long-argued alternative becomes relevant. A Paris-based international lawyer who has argued in this journal since the 1990s for a framework he calls “Two States, One Holy Land,” Whitbeck does not abolish statehood so much as soften the separation it implies: two states bound confederally, the land left undivided, and Jerusalem held under joint sovereignty as a shared and undivided capital. His is a refusal of hard partition rather than of statehood itself, and it points away from the sealed-entity logic that Dayton entrenched. Dayton’s first lesson for Israel-Palestine is thus cautionary: territorial entities are a tool for places where the cartography of separation already exists, and the harder a settlement leans on them, the more it codifies displacement as a precondition for peace.
The High Representative
If the entity structure was Dayton’s skeleton, the Office of the High Representative became its nervous system. Conceived in Annex 10 as a weak coordinator of civilian implementation, the office acquired sweeping authority at the 1997 Bonn conference: the power to impose laws and to dismiss elected officials who obstructed the peace. For two decades High Representatives used these “Bonn powers” liberally, and critics such as Knaus and Martin described the result as a “European Raj,” a protectorate in which foreign officials set the agenda and treated a sovereign state as an unruly minor endlessly prepared for a self-government it was never allowed to reach. The present crisis is the clearest illustration the office has ever produced. High Representative Christian Schmidt blocked Republika Srpska legislation, Dodik defied him and was convicted for it, Schmidt then suspended public funding to Dodik’s party, and the entire confrontation ended with the overseer himself resigning under disputed legitimacy. The former High Representative Wolfgang Petritsch, surveying this in late 2025, concluded that the office should be wound down and its oversight role passed to the European Union, on the ground that the deeper problem is less any individual High Representative than the sheer longevity of the office itself.
For Israel-Palestine the relevant question is brutally simple: who would be the High Representative, and would anyone obey him? Bosnia’s office functioned, to the degree it did, because three conditions held that are absent here. There was a single dominant external guarantor in the United States, a defeated party in Republika Srpska, and a broad international consensus behind the project. None of these exists in the Holy Land, where the external guarantors do not agree among themselves, neither party has been militarily resolved, and neither would accept an outside official empowered to remove its elected leaders. The very nature of the parties imports a symmetry that does not exist: a High Representative here would not be holding apart two spent belligerents but administering a relationship between an occupying power and a population without sovereignty, which is nearer the colonial protectorate Knaus and Martin warned Bosnia was becoming than a neutral peace office. The deeper lesson runs beneath the logistics. The Bonn powers are evidence that consociational power-sharing in a post-war society does not run on its own. It had to be driven by a quasi-colonial executive, and thirty years on it has produced dependency rather than a self-sustaining democracy. Any framework for Israel-Palestine that rests on an international administrator inherits both that dependency and the legitimacy deficit that the non-recognition of Schmidt has now exposed for all to see.
The office also carries the lesson the Holy Land least wants to hear, about the half-life of the word interim. A transitional instrument is still issuing binding decrees thirty years on, because nothing was built to make it end. Oslo has run this experiment once already: its five-year interim period, and the division of the West Bank into Areas A, B and C pending a final status that never arrived, was meant to expire in 1999, and a quarter-century later the provisional arrangement simply is the arrangement. Bosnia’s warning is that interim structures do not decay on their own. They accrue constituencies, foreign and domestic, that come to prefer the freeze to the risk of resolution, and any framework that does not write its own expiry into law will be administering the same limbo at its own thirtieth anniversary.
Ethnically Defined Representation
Dayton wrote ethnicity into the architecture of the state. The presidency is a trio, one Bosniak, one Croat, one Serb, the first two elected from the Federation and the third from Republika Srpska; the House of Peoples is composed on the same ethnic parity; and a vital-national-interest veto allows each group to block legislation it deems threatening. The predictable consequence was paralysis, an immense and immobile apparatus in which the system left almost no room for non-ethnic politics and gave no incentive to debates that could not be translated into ethnic ones. Its most damning monument is the European Court of Human Rights ruling in Sejdić and Finci in 2009, which found that barring a Jewish and a Roma citizen from the presidency and the upper house was unlawful discrimination. Sixteen years later the judgment remains unimplemented, because the groups the constitution privileges are the same groups whose consent would be required to reform it.
This is at once the most directly relevant mechanism for Israel-Palestine and the most cautionary. Reserving office by ascriptive group identity entrenches the very cleavage it is meant to manage. A formula that fixed “Jewish” and “Palestinian” quotas into the structure of the state would make identity the currency of citizenship, marginalise everyone who does not fit the categories, the secular, the mixed, the Palestinian citizens of Israel who already occupy an uneasy constitutional space, and reproduce the same veto-driven deadlock. It would also assume what the occupation denies, a single body of citizens to apportion seats among, when the population to be represented is in fact split between those who vote in Israeli elections and those who, under occupation, hold no franchise at all. The prior question is not how to share power but who counts as a member of the polity, and Bosnia, where membership was at least settled, cannot answer it. There is one usable insight here. In conditions of total mistrust, the principles of proportional inclusion and a mutual veto on existential questions have genuine appeal, and may be unavoidable as a transitional device. The challenge, which Dayton failed and which Whitbeck’s parity-of-peoples logic tries to meet differently, is to deliver that reassurance without making ethnicity constitutive of the state itself.
The Constitutional Court
The subtlest of Dayton’s mechanisms is also the one whose lesson travels furthest. Bosnia’s constitution is not a domestic document at all; it is Annex 4 of an international treaty, drafted abroad and never ratified by any Bosnian institution. When the Constitutional Court was asked in 1997 to review the framework’s validity, it answered that it could not, since the court itself existed only by virtue of that constitution and could not sit in judgment of its own source. The foundational law was therefore effectively unreformable from within, what Marković calls an octroi constitution, imposed rather than authored. The result is a constitution that most citizens regard as a foreigner’s text, preserving the country’s borders while never generating the social cohesion or sense of ownership that a constitution is supposed to embody.
For Israel-Palestine the ownership problem is the sharpest transferable lesson of all. A settlement drafted at an airbase by external powers and annexed to a ceasefire never acquired domestic legitimacy, and the longer and deeper the conflict, the more essential that sense of authorship becomes. A framework imposed from outside, or merely perceived as imposed, inherits this defect at birth. The court’s structural subordination to the treaty carries a second warning about rigidity. In Bosnia rigidity became the trap: a settlement locked so tightly that it cannot evolve calcifies into the status quo it was meant to transcend. Any durable arrangement in the Holy Land needs an internal path to amendment and a credible claim to local authorship, or it will harden into the next grievance.
Against Caution: Five Lessons
An examination of warnings is not a peace plan, and an anniversary is a poor occasion for counsels of caution alone. The status quo is not a low-risk baseline against which every proposal looks dangerous. It is itself the most dangerous option on the table, a slow-motion catastrophe that manufactures the next war while everyone waits for conditions to ripen. In that spirit the following five lessons are offered deliberately, and some of them carry real risk. They share one premise: that the lesson of Dayton is not which of its parts to copy, but that a settlement must be built for the morning after the guns fall silent, and built to evolve.
The first lesson is to transfer the plan that failed rather than the one that succeeded. Dayton is remembered because it worked, and the Vance-Owen peace plan, the joint UN and European Community proposal drawn up by Cyrus Vance and Lord Owen, because it did not. Yet on its merits Vance-Owen was the more ambitious design: multi-ethnic provinces, a deliberate denial of contiguous territory to the Serbs, and a structure that made interdependence the only available future. It failed for want of enforcement, not for want of vision. Israel-Palestine, where the populations are interwoven rather than sorted, fits the Vance-Owen logic of interlocking, non-contiguous, multi-community units under a thin shared roof far better than it fits Dayton’s clean dividing line. The proposal is not to revive its map but to recover its principle, that in a genuinely mixed land, separation is the fantasy and interdependence the only realism.
The second lesson is to federate functions rather than territory. Where a line cannot honestly be drawn on the ground, it can be drawn around people and competences instead. Belgium already distinguishes territorial regions from non-territorial communities whose authority follows the citizen wherever she lives. A single shared space could carry the same logic: parallel civic and personal-law systems for those who want them, and joint institutions for everything that cannot be divided, among them water, spectrum, antiquities, the holy sites, and the economy itself. The risk must be named rather than buried, as parallel systems can ossify into the very separateness they were meant to soften, and the design works only if it is bound from the outset to a single enforceable bill of rights and one common court with jurisdiction over both communities.
The third lesson is to build an international guarantor designed to dissolve. Dayton’s overseer had no expiry and no exit, and three decades later it is still there, breeding the very dependency that leaves Bosnian politics unable to stand on its own. The answer is not to refuse external help, which is what stopped the killing, but to legislate its decay. A guarantor whose powers diminish on a fixed schedule and transfer to jointly staffed domestic institutions as capacity benchmarks are met, with those powers lapsing by default if the transfer stalls rather than persisting by inertia, is an overseer that plans its own funeral instead of becoming the disease it was sent to cure.
The fourth lesson is to make representation centripetal rather than consociational. The deadlock of Bosnia’s ethnic quotas and vetoes, and the standing scandal of Sejdić and Finci, the Roma and Jewish citizens barred from high office as non-members of the three constituent peoples, are arguments against Arend Lijphart’s grand-coalition model, set out in his Patterns of Democracy, not for it. The alternative is to engineer electoral rules that reward politicians for reaching across the divide: vote-pooling, distribution requirements that oblige a candidate for shared office to earn meaningful support within the other community, and a chamber organised around cities rather than peoples. The aim is to make moderation electorally profitable rather than merely constitutionally mandatory. Centripetalism has a mixed record and can fail where polarisation is absolute, which is the honest risk, but the consociational record in conditions resembling the Holy Land is worse.
The fifth lesson is to write a constitution that is deliberately incomplete and dually ratified. Dayton’s founding text was imposed at an airbase, owned by no one, and effectively unreformable from within. Its opposite is a thin shared charter, ratified by referendum on both sides so that authorship is local and the document is felt as one’s own, which intentionally leaves the hardest questions, final sovereignty over Jerusalem and the modalities of return, to scheduled future negotiation under binding and pre-agreed procedures. This is the method of the incompletely theorised agreement, settling what can be settled now and building the road to the rest. The danger is that deferral becomes permanent evasion, and it can be contained only by hard deadlines and a default to binding arbitration when those deadlines pass.
What the Anniversary Actually Teaches
Set against the question of what realistic alternatives to war exist, Dayton at thirty teaches mostly by inversion. It succeeded at peace-making and failed at state-building, and those are different crafts. It ended a war, a real and humane achievement won partly through Holbrooke’s instinct that what did not get done at Dayton would not get done later. But it then froze a captured, externally administered state that cannot reform itself and whose founding compromise made ethnicity the only legitimate political identity. Each of its mechanisms, read closely, points to its own correction: territorial separation toward functional federation, the unaccountable overseer toward the self-dissolving guarantor, ethnic quotas toward centripetal incentives, the imposed and frozen constitution toward the locally authored and deliberately open one.
That is the deeper reason a framework like Whitbeck’s, which keeps the land undivided while dividing sovereignty between two states and holding Jerusalem in common, points away from the sealed entities of the Dayton map. Dayton is invoked for Israel-Palestine because it is the famous case of a war that was stopped, and the anniversary is a reminder that stopping a war and building a country that can govern and reform itself are not the same task. The courage on display in 1995 was the courage to put everything on paper while leverage lasted. The courage the next thirty years will demand is the courage to build something better than Dayton, and to accept that in a conflict this old the cautious option and the catastrophic one have quietly become the same thing.