(This commentary appears in two parts. Part one examines the silences in the founding statement; part two takes up the distribution of power inside the alliance)
Part One: The Silences in the Statement
By Worku Aberra
Seven armed organizations announced on 20 September 2026 that they had formed the Ethiopian Peoples’ Forces Alliance for Survival, known in short as the Alliance for Survival: the Amhara Fano National Movement, the Oromo Liberation Army, the Tigray People’s Liberation Front, the Ogaden National Liberation Front, the Afar Revolutionary Democratic Unity Front, the Benishangul Peoples’ Liberation Movement, and the Gumuz Peoples’ Democratic Movement. Two of the seven have since disputed the claim. The Gumuz movement denies membership and says it has no armed wing; one wing of the Ogaden front denies membership while another confirms it (Raleigh and Getachew Birru 2026). An arrangement uncertain of its own membership is in no position to settle the larger questions it left open.
The signatories declared two objectives: removal of the current government and establishment of an inclusive transitional government under a transitional charter. The achievement is real. A single opposition bloc raises the price of divide and rule, a strategy Abiy Ahmed’s government has used to good effect: an incumbent who faces separate opponents concedes to some, represses others, and never confronts concerted pressure. Mediators also gain a single interlocutor, one whose agreement would bind all its members. Multi-regional composition, spanning Amhara, Oromia, Tigray, Afar, Somali, and Benishangul-Gumuz, denies the government its preferred characterization of each insurgency as one group’s quarrel with the state.
What the Name Concedes
The alliance’s name requires a close reading, since both of its operative terms were chosen from a set of available alternatives. Students of armed politics distinguish three forms of cooperation. An alliance is a formal agreement among independent actors to coordinate the pursuit of shared objectives, and each member keeps its own forces, leadership, and aims. A front subordinates its members to a joint command under a common programme. A coalition draws in a wider membership than either: armed movements alongside political organizations, civic associations, and professional bodies.
The distinction matters, since the form cooperation takes shapes how a partnership behaves once power shifts among its members (Christia 2012). The signatories adopted the least integrated of the three forms, and their choice of name reports it accurately. One member had said as much before the announcement: on 17 September the OLF-OLA High Command described the emerging arrangement as a “procedural consensus,” denied that any participating group held a mandate to settle Ethiopia’s political questions, and stated that the OLA was neither involved in nor a component of the alliance (OLF-OLA High Command 2026; Addis Standard 2026b). Six days after the signing, ACLED’s regional analyst reported that the OLA had joined to see the government removed and assumed no commitment to joint operations. Only three members were then fighting federal forces: the TPLF, Fano, and the Afar Ugugumo (Raleigh and Getachew Birru 2026).
The second term is less exact. Survival carries at least five meanings: physical survival of civilian populations; organizational survival of the signatories under military pressure; survival of each ethnic collectivity as a political unit entitled to self-rule; survival of the state within its present borders; and survival of a constitutional order, whether the existing one or its successor. The statement selects none of them. Only the third unites the signatories: each movement claims to defend the survival of a particular people.
The fourth meaning appears nowhere in the statement. Its authors speak of “the security of the country” and never of the survival of the state, an omission that follows from the membership: two signatories hold doctrines that permit exit, and neither could endorse the state’s survival as an object of the struggle. The second meaning explains the timing best, and the TPLF’s case is the clearest. Every signatory faced military or political pressure through 2026, and the organization in Mekelle faced the most: a rival Tigrayan leadership operating from Addis Ababa, an armed Tigrayan splinter on its southern border, and clashes in January between its forces and federal troops over a district it claims.
Students of alliance formation treat threat to an organization’s own existence as the standard trigger for alliance among rivals (Walt 1987; Christia 2012). Refusal to choose among these meanings performs the same function as the silence on national unity: each constituency reads its own, and no member pays in advance for a concession it may never have to make.
Whose Survival, and Against What
Supporters of the alliance press the first of the five meanings furthest: they hold that the government threatens the physical survival of Ethiopians as a whole. The claim merits examination on its own terms. Two questions follow: what such a claim requires in evidence, and whether the record in several regions establishes it for the whole population.
Can a government, even a dictatorial one, pose an existential threat to the people it governs? The historical record answers in the affirmative. Germany under the Nazi regime from 1941 to 1945, Cambodia under the Khmer Rouge from 1975 to 1979, and Rwanda under the Hutu Power government in 1994 each show a state apparatus turned against populations it claimed to govern. Regimes that undertake such campaigns share identifiable features: an ideology that marks a category of people for removal, an administrative machinery built to that purpose, and a leadership willing to bear the costs of destruction rather than the costs of accommodation.
Such regimes are rare, and the incentives explain why. A government that destroys its own population destroys its tax base, its conscripts, and the labour that sustains the economy it extracts from. Mass killing also invites intervention, sanctions, and prosecution, and it forecloses the bargains that keep rulers in office. Regimes that rule through a narrow coalition rely on selective repression for that reason: they punish the groups that threaten them and court the groups they need, since selectivity costs less and serves the purpose better than annihilation. A regime that threatened everyone would be acting against its own interest.
Some commentators nonetheless describe the government as an existential threat to Ethiopians as a whole. Their reasoning runs from the record in Tigray, Amhara, and Oromia to a conclusion about the country, and those three regions hold most of its population. The step is worth examining closely. Take a regime that does threaten the survival of one group: it targets that group’s leadership, expels its members from a contested district, or starves its region of food, medicine, and other essentials. Measures of that kind work by distinguishing the group from everyone else. Selectivity is what makes them effective, and selectivity is what prevents them from generalizing. What is true of several parts is therefore not true of the whole, and the inference commits the fallacy of composition.
A second objection concerns the premises rather than the reasoning. A regime at war in three regions is fighting three armed movements rather than three peoples, and counterinsurgency differs from a campaign of destruction, whatever the suffering it inflicts on civilians. Neither objection denies that the count matters. A government that turned on most of its populations at once would come close to the universal case, and the claim would then rest on direct evidence rather than on an inference.
The Missing Commitment, and What Followed
The statement affirms nothing about Ethiopia’s national unity and territorial integrity, and the omission is not an oversight. Two of the signatories invoke Article 39 of the 1995 constitution. The TPLF led the drafting of the document it now relies on, and the article’s guarantee of self-determination up to secession supplies each of the two with a claim the alliance must accommodate. The TPLF keeps the right in reserve and has never renounced it. The OLA fights for Oromo self-determination and leaves the destination unstated. In the declaration the article appears again, unnamed: decisive questions concerning the peoples and the country are to be “resolved with the direct participation of the peoples” (Alliance for Survival 2026). That formulation reads as democratic consultation to one constituency and as the constitutional procedure to another. A pledge of territorial integrity would have obliged both signatories to abandon a founding claim, and neither had reason to pay that price.
The war that began three days after the announcement bears on the question left open. On 23 September, Tigrayan forces seized the airports at Mekelle, Aksum, and Shire, and opened fronts along the Amhara and Afar borders (Raleigh and Getachew Birru 2026). The TPLF’s statement describes a “genocidal invasion” and a “defensive battle” (National Regional Government of Tigray 2026); Getachew Reda, an adviser to the Prime Minister and a former TPLF leader, describes an all-out Tigrayan offensive (Al Jazeera 2026b).
A charge of that gravity points toward several remedies. Its authors might seek prosecution of the perpetrators, international protection for the population, constitutional guarantees against a repetition, or secession. The last gains the most from the language, since a leadership that describes the state as an existential aggressor will struggle to justify remaining within it. Support for Tigrayan independence has meanwhile become an organized demand. The Voice for Independent Tigray, a diaspora group, made independence its central demand in a declaration of 31 May 2025, and Tigrayan commentators arguing on both sides describe the question as newly mainstream and sharply polarizing (Fitwi 2025). A second war fought in these terms serves its advocates well.
The Amhara Fano National Movement’s place in the arrangement is the harder puzzle. Three days after the signing, its units fought beside Tigrayan forces along the Kobo–Tumuga axis, a corridor both groups claim (Raleigh and Getachew Birru 2026). Fano took up arms over what its leaders describe as the persecution of the Amhara people and the threat to their existence. Its programme includes opposition to the country’s break-up, and its partner’s constitutional position permits precisely that outcome. Fano’s fighters are therefore clearing a path for an outcome their movement exists to prevent. Analysts reading the alliance as a partnership of diffuse and sometimes contradictory interests have identified the same weakness (Ross 2026; CNN 2026). Nothing in the statement provides for the day the enemy is gone.
Events are moving quickly, and their direction remains open. Commentary so far has settled on one question above the rest (Raleigh and Getachew Birru 2026; Ross 2026): can the alliance defeat Abiy Ahmed’s government? A second question matters more and has drawn less attention: what would follow if it did? An alliance able to remove a government may lack the capacity to replace one. The TPLF already governs a region, collects revenue, and commands an army; the others hold ground and answer to no common authority. Victory would leave every member in possession of what it already holds. A centre binds the regions to one another, settles what they cannot settle between them, and speaks for them abroad. The removal of the government could leave the country without one.
A framework for reading the alliance
Three findings from the study of armed conflict organize the argument that follows: how partners choose each other, what binds them to their promises, and how internal structure shapes a group’s prospects.
Armed groups weigh power calculations more heavily than ideological affinity in choosing their partners. Each asks which combination gives it the best position at the end of the fighting, and the answer changes as the war proceeds, so yesterday’s ally could become today’s adversary when the balance moves (Riker 1962; Christia 2012). Cooperation among former enemies therefore requires no ideological convergence, and it settles nothing about the order that follows victory, since the calculation that produced the alliance will be made again the moment the common enemy falls.
What then holds partners to the terms they agreed? Very little, on the evidence. Promises made under mutual dependence cease to bind once the dependence ends; anticipation of that shift can provoke preventive violence, since a partner who expects betrayal has reason to strike first (Fearon 1995; Powell 2006). The remedy, where one exists, lies outside the partnership. Settlements between warring parties survive when a third party guarantees them, since disarming leaves each side exposed to the other (Walter 1997, 2002). Partners in an alliance face the same exposure before any settlement is reached, and a guarantor would serve a different purpose for them: preventing a conflict rather than ending one. What decides the outcome is enforceability rather than good faith.
A movement’s internal structure shapes its prospects as well. Krause finds that a movement under a hegemonic organization achieves its goals more often than a fragmented one, and hegemony here denotes political command rather than superior force: the capacity to speak for the movement, to discipline rivals, and to deliver what a bargain requires (Krause 2017). An assumption underlies the finding. Krause’s movements contest the leadership of a struggle whose object all members accept, and partners who differ over the object have no reason to accept any leader. Fragmentation then produces the opposite of hegemony. It multiplies veto players, invites the incumbent to concede to some factions and repress others, and raises the probability that one faction defects for a separate settlement (Cunningham 2011, 2013; Pearlman 2008/09).
These findings yield a proposition about an alliance with three characteristics. The first is incompatible purposes: each member exists to pursue an end the others were founded to resist, so none can command the rest. The second is the absence of a guarantor: no external power stands behind what they promise one another. The third is indivisibility: the territory they contest admits no partition. Such an alliance can remove a government and fail to replace it. Victory would leave every member holding a region, an army, and a claim the others reject, and nothing above them able to decide between the claims.
Programmes that point in opposite directions
The Alliance for Survival has the first of those three characteristics. Its three largest signatories, the Amhara Fano National Movement, the TPLF, and the OLA, pursue ends that cannot all be realized: two hold programmes that permit departure from the Ethiopian state, and the third exists to prevent it.
The TPLF’s position rests on a right it has never renounced. Article 39 remains available, and the organization has said nothing to close that door. Its conduct meanwhile points away from the federal framework: it removed the Tadesse Worede interim administration in April, installed Debretsion Gebremichael as president in May, and treats the Pretoria arrangement as void. Taken together, the steps describe a regional government operating as though federal authority had lapsed. Whether that is a rehearsal for secession or a negotiating position, the organization has left both readings open, and the ambiguity serves it as the statement’s ambiguity serves the alliance.
The OLA’s manifesto binds the movement to the Oromo people’s right to freely determine their political status, a formulation it grounds in the international covenants and one that admits independence among its outcomes (OLF-OLA High Command 2023). Its current demands point inward: Afaan Oromoo as a federal working language, recognition of Finfinnee as an integral part of Oromia, and an inclusive settlement among all stakeholders. Supporters of the movement press further and advocate outright independence, and the manifesto leaves them the ground to do so.
Fano has pledged itself against the country’s break-up (Lema 2026), and its territorial claims depend on the state remaining whole. Welkait, Tselemt, and Raya are disputed between two regions of a single federation, and a federal government can settle such a dispute through a referendum, a boundary commission, or a court. Tigrayan secession would close every one of those routes and leave conquest as the only remaining one, and a war between states at that.
The alliance’s statement accommodates all three positions by saying nothing. It borrows the constitution’s language of “nations, nationalities and peoples,” defers the question of the future order to a dialogue it does not describe, and promises to settle differences according to “the freely expressed will of our peoples” (Alliance for Survival 2026). Those terms follow from the membership: every signatory organizes around a single ethnic or regional constituency, and none claims to represent Ethiopians as a whole. An alliance of that composition has no member able to speak for the state it proposes to reconstitute. The TPLF and the OLA needed language that left Article 39 intact; Fano, whose programme calls for revising the constitution, signed a text written in its terms.
The signatures were bought with silence. Every question the statement postponed concerns the state rather than the government: whether Ethiopia remains one country, where its internal boundaries run, and which constitution governs whoever wins. The boundary question alone reaches contested areas across the north and west, and the capital sits outside its reach altogether. Whether the postponement matters depends on a further question the statement leaves open. Programmes that cannot be reconciled matter only if their holders can impose them, and the capacity to impose belongs to whichever partner emerges strongest. The second part of this commentary asks which partner that might be, and takes up the two characteristics left aside here: whether anyone stands behind the partners’ promises, and whether the prize they contest can be divided at all.
Editor’s Note: Views in the article do not necessarily reflect the views of borkena.com
__
Join our Telegram Channel: t.me/borkena
Like borkena on Facebook
To submit a press release, send the submission to info@borkena.com
Add your business to Ethiopian Business Listing / Ethiopian Business Directory
Join the conversation. Follow us on X (formerly Twitter) @zborkena to get the latest Ethiopian news updates regularly.
References
Addis Standard. 2026a. “Tigray Interim President Says TPLF Move to Reinstate Pre-War Council ‘Nullifies’ Pretoria Agreement.” 21 April. https://addisstandard.com/tigray-interim-president-says-tplf-move-to-reinstate-pre-war-council-nullifies-pretoria-agreement/
Addis Standard. 2026b. “OLA Describes Emerging Alliance among Opposition Armed Groups as Procedural Consensus, Not Political Agreement.” 18 September. https://allafrica.com/stories/202609180258.html
Al Jazeera. 2026a. “After Pretoria Peace Deal, Old Enemies in Ethiopia’s Tigray Find New Ground.” 5 September. https://www.aljazeera.com/news/2026/9/5/after-pretoria-peace-deal-old-enemies-in-ethiopias-tigray-find-new-ground
Al Jazeera. 2026b. “Tigray Forces Launch Offensive against Ethiopian Government.” 23 September. https://www.aljazeera.com/news/2026/9/23/ethiopian-airlines-suspends-flights-to-three-northern-tigray-airports
Christia, Fotini. 2012. Alliance Formation in Civil Wars. Cambridge: Cambridge University Press.
CNN. 2026. “War Has Returned to Ethiopia’s Tigray, and Former Enemies Are Now Fighting Together. What Changed?” 25 September. https://edition.cnn.com/2026/09/25/africa/war-returns-to-ethiopias-tigray-intl
Cunningham, Kathleen Gallagher. 2011. “Divide and Conquer or Divide and Concede: How Do States Respond to Internally Divided Separatists?” American Political Science Review 105 (2).
Cunningham, Kathleen Gallagher. 2013. “Actor Fragmentation and Civil War Bargaining: How Internal Divisions Generate Civil Conflict.” American Journal of Political Science 57 (3).
Ethiopian Peoples’ Forces Alliance for Survival. 2026. Founding statement, 20 September. English text as circulated.
Fearon, James D. 1995. “Rationalist Explanations for War.” International Organization 49 (3).
Fitwi, Abrar. 2025. “Tigray Needs Democracy before Secession.” Ethiopia Insight, 22 July. https://www.ethiopia-insight.com/2025/07/22/tigray-needs-democracy-before-secession/
Krause, Peter. 2017. Rebel Power: Why National Movements Compete, Fight, and Win. Ithaca: Cornell University Press.
Lema, Luis. 2026. “Zemene Kassie, chef de la milice Fano: «Nous luttons contre la désintégration de l’Ethiopie».” Le Temps, 16 August. https://www.letemps.ch/monde/afrique/zemene-kassie-chef-de-la-milice-fano-nous-luttons-contre-la-desintegration-de-l-ethiopie
National Regional Government of Tigray. 2026. Statement, 23 September. Translation at https://martinplaut.com/2026/09/23/tigray-authorities-ethiopian-government-has-invaded-us-we-are-fighting-back/
OLF-OLA High Command. 2023. A Brief Political Manifesto: From Armed Struggle to the Prospect for Peace. January. https://www.fnnmedia.org/blog/2023/1/24/brief-political-manifesto-olf-ola
OLF-OLA High Command. 2026. Press release on the emerging alliance, 17 September.
Pearlman, Wendy. 2008/09. “Spoiling Inside and Out: Internal Political Contestation and the Middle East Peace Process.” International Security 33 (3).
Powell, Robert. 2006. “War as a Commitment Problem.” International Organization 60 (1).
Raleigh, Clionadh, and Jalale Getachew Birru. 2026. “Ethiopia Conflict: Expert Insights.” ACLED, updates of 23 and 26 September. https://acleddata.com/expert-comment/ethiopia-conflict-expert-insights
Riker, William H. 1962. The Theory of Political Coalitions. New Haven: Yale University Press.
Ross, Aaron. 2026. “Flush with Allies, Tigray’s Leaders Face Internal Discord in New Ethiopia War.” Reuters, 24 September. https://www.reuters.com/world/africa/flush-with-allies-tigrays-leaders-face-internal-discord-new-ethiopia-war-2026-09-24/
Walt, Stephen M. 1987. The Origins of Alliances. Ithaca: Cornell University Press.
Walter, Barbara F. 1997. “The Critical Barrier to Civil War Settlement.” International Organization 51 (3).
Walter, Barbara F. 2002. Committing to Peace: The Successful Settlement of Civil Wars. Princeton: Princeton University Press.
Worku Aberra is a scholar in residence in the Economics Department at Dawson College in Montreal.