Editorial perspective · Part 2 of 28
The Price Without the Return · I — Framing
The Identity Question: Why Nations Need One and What Happens When They Lose It
A nation is a political community whose members recognise each other as members of the same political community. This is a functional definition, not a romantic one.
An argument by Zef Telahun
This is an editorial perspective — signed opinion, not the site's neutral analysis. Factual claims are footnoted; the synthesis, emphasis, and judgement are the author's.
Thesis
A nation is a political community whose members recognise each other as members of the same political community. This is a functional definition, not a romantic one. It does not require that members share ethnicity, religion, language, or history — many successful nations contain populations that share none of these fully. What it requires is that, at the level of national political question, members regard themselves as one people making one set of decisions about one shared future. Where this recognition exists, the state has a nation to govern. Where it does not exist, the state has a collection of populations to hold together by force, negotiation, or accident, and the political-economic work that nations can undertake (industrial coordination, universal education, collective defense, redistributive taxation, integrated infrastructure) becomes systematically harder.
Ethiopia in 2026 does not have a nation in this functional sense. Ethiopia has a state that governs a territory in which multiple large populations do not, at the level of national political question, regard themselves as one political community. The Tigrayan population has, since November 2020, been in a relationship with the federal government that ranges from open military conflict to uneasy separation. The Amhara population has, since April 2023, been in armed conflict with the federal government across large parts of the region. Substantial parts of the Oromo population support the OLA insurgency; substantial parts of the Somali population maintain political distance from federal authority; the various southern nationalities have used the 1995 Constitution’s Article 39 to subdivide the SNNPR into four successor regions, with more subdivisions being requested. The population that unequivocally identifies itself, at the level of national political question, as first Ethiopian and second whatever else, is a minority within the state Ethiopia claims to govern.
The article’s argument is that this is not a description of a country in temporary difficulty. It is the description of a country that has never constructed the identity a modern state requires, and whose absence of that identity is now the operative obstacle to every other developmental problem. The Asian comparators the series examines built pan-national identities that could hold their populations together as one political community. Ethiopia’s regimes have not. This is what the identity failure means, and it is the deepest of the three failures this series is examining.
What identity means, functionally
The word “identity” carries a lot of intellectual freight — cultural, psychological, ethnic, religious, historical. This series uses it in one specific sense: the sense in which a population’s members answer, when asked at the level of national political question, “we are one people.” The definition is functional and behavioural, not essentialist and cultural. It does not require that people love their fellow citizens or feel warmly toward them or share their tastes. It requires that when the state raises taxes, they consent to pay because the taxes will fund shared purposes. That when the state calls for national defense, they respond because the defense is theirs. That when the state redistributes resources from wealthier to poorer regions, the wealthier regions accept the redistribution because the poorer regions are one of us. That when a national election is held, the losers accept the outcome because the winners, though they may be from other regions or groups, are legitimate representatives of one shared political community.
This functional definition of national identity is what political theorists like Ernest Gellner and Benedict Anderson have analysed at length. Gellner’s Nations and Nationalism (1983) argues that industrialisation requires the standardisation of language, education, and identity across a population so that individuals can move fluidly between economic roles.1 Anderson’s Imagined Communities (1983) argues that modern nations are “imagined” — constructed in the political imagination of their members — through print media, standardised education, and shared public rituals.2 Both frameworks converge on the empirical observation that successful modern states have constructed shared identities among populations that were previously fragmented, and that the construction was a deliberate political project sustained across generations, not the natural expression of pre-existing cultural unity.
Applied to Ethiopia, both frameworks are diagnostic. The Ethiopian state has not, since 1974, sustained the deliberate political project of identity construction. The population is not, in Gellner’s sense, culturally standardised in ways that permit fluid economic mobility across the country. The population is not, in Anderson’s sense, imagining itself as one political community through shared media, shared education, and shared rituals — it is imagining itself, in significant part, as multiple communities whose relationship to each other is contested.
What identity produces
Nations that have constructed shared identity — Japan, South Korea, Taiwan, Singapore, China at various stages, Malaysia imperfectly, Indonesia through Pancasila, Turkey through Kemalism, France through the Republic, Germany after 1945 through constitutional patriotism, India (contested but sustained) through the Congress-era nationalism — receive from their identity a set of political-economic capacities that populations without shared identity cannot exercise.
Coordination of industrial development. Industrialisation requires labour mobility, capital allocation across sectors, and infrastructure investment that connects regions. Populations that regard themselves as one nation accept the reallocation because it serves shared purposes. Populations that regard themselves as multiple nations resist the reallocation, because their region’s resources are being taken to serve another region’s development. Ethiopia’s post-2005 large-scale land leasing programme, the GERD financing that drew resources disproportionately from certain regions, and the industrial-park siting decisions have all produced local resistance that a shared identity would have moderated.
Universal education in a standardised curriculum. Modern economies require workforces educated in shared bodies of knowledge — mathematics, sciences, national language, historical narrative — that permit individuals to move between regions and sectors. Populations that share national identity accept the standardisation because the shared curriculum expresses their shared community. Populations that do not accept the standardisation as cultural imposition; the curriculum becomes contested and, in the extreme, is replaced by regional variants that reproduce the pre-existing fragmentation. Ethiopia’s post-1991 regional-language education policy was operationally correct in its recognition of Ethiopian linguistic diversity, but the accompanying construction of a shared national curriculum that could integrate the regional variants was not undertaken.
Universal military service and collective defense. Nations that share identity can conscript their members into military service; the members serve because the defense is theirs. Nations that do not share identity cannot conscript reliably; the members serve, if they serve at all, on the basis of coercion or particular incentive. The Amhara Special Forces dissolution of April 2023 illustrates the point. The federal decision to integrate regional security forces into a unified federal command was a nation-building move in structure; it presupposed that Amhara personnel would serve the federal state on the same terms as Oromo or Somali personnel. The presupposition was not accurate to the political reality, and the resistance produced the Fano insurgency that has continued for three years.
Redistributive taxation and revenue-sharing. All modern states redistribute resources from wealthier regions and populations to poorer ones through taxation and public expenditure. Nations that share identity accept the redistribution as an expression of shared community. Nations that do not share identity treat the redistribution as extraction. The Ethiopian federal revenue-sharing system, established under Solomon Negussie’s fiscal-federalism framework in the mid-1990s, has operated throughout the post-1995 period but has been chronically contested by regional actors who perceive the sharing as unfair to their region. The perception is not entirely unfounded — the specific formulas have been subject to political manipulation — but the deeper issue is that no shared identity underwrites the sharing.
Democratic legitimacy of election outcomes. Nations that share identity accept electoral outcomes even when their preferred candidate loses, because the winner is a legitimate representative of the shared community. Nations that do not share identity contest electoral outcomes when their region’s candidate loses, because the winner represents another community. The 2005 Ethiopian election produced opposition wins in Addis Ababa and the Amhara region that the EPRDF suppressed; the 2021 and 2026 elections produced outcomes that opposition parties largely boycotted or rejected; the 2020 postponement produced the constitutional crisis that led to the Tigray war. In every case, the electoral outcome could not be accepted by the losing side as an expression of shared national decision.
The list could be extended. The general pattern is that shared national identity produces the political-economic capacities of the modern state. Its absence forecloses those capacities. This is why identity construction is not an optional cultural project. It is a structural precondition for the state’s ability to do the work modern states are supposed to do.
What Ethiopia had
The Ethiopian question is not whether the country has ever had an identity. Ethiopia has had identities. The question is whether those identities have been of the kind and scope that a modern multi-ethnic state requires. This series argues they have not.
The imperial-era identity — what one might call the Solomonic-Christian-Amharic imperial package — was the identity around which the Ethiopian state was organised from the late-medieval period through 1974. This identity had substantial content. It included the claim of Solomonic descent for the ruling dynasty (the Kebra Nagast tradition), the Ethiopian Orthodox Church’s central role in political legitimation, the Amharic language as the vehicle of imperial administration and high culture, the Ge’ez liturgical tradition, and the Aksumite historical inheritance as the state’s origin story. It gave Ethiopia a distinctive place in the world: the African Christian empire that had preserved its independence through the colonial period, that was recognised at the League of Nations, that hosted the Organisation of African Unity in Addis Ababa.
For the populations who were incorporated into this identity — primarily the Amhara, and to significant degrees the Tigrayan, Gurage, and some other groups — the imperial identity was theirs. They participated in it as their own tradition, contributed to its development, and derived from it a real sense of Ethiopian-ness that has continued to inform Ethiopian nationalist politics through 2026.
For other populations, the identity was not theirs. The Oromo populations, incorporated into the imperial state through the conquests of Menelik II in the late nineteenth century, did not participate in the Solomonic-Christian-Amharic package as their own; they experienced it as the identity of the conquering power, imposed on their own oromo cultural, linguistic, and religious traditions. The Muslim populations of the empire — a substantial portion including the Somali, Afar, and large parts of the Oromo — did not participate in the Christian dimension of the imperial identity. The southern nationalities incorporated in the same nineteenth-century expansions had their own languages, religions, and political traditions displaced by the imperial framework. The Tigrinya-speaking populations of Eritrea, incorporated into Ethiopia after the 1952 federation and the 1962 dissolution, had experienced Italian colonial administration for fifty years and had developed their own distinct political culture that the Ethiopian imperial framework did not accommodate.
The imperial-era Ethiopian identity was, in short, real but partial. It was one identity among several within the territory the state governed, and it was operationally the identity of the ruling class and the culturally-privileged populations rather than of all Ethiopian subjects. This partiality is what the imperial framework did not address across four decades between the 1931 Constitution and the 1974 revolution. Haile Selassie’s modernising reforms produced the form of a modern state (constitution, parliament, cabinet, ministries, universities) but did not produce the substantive identity construction that would have made the state the political expression of one nation. The 1974 revolution’s most immediate political fact was that the imperial identity, while contested, was the only identity Ethiopia had available at national scale; when the revolution destroyed the imperial framework, it destroyed the only national identity Ethiopia possessed.
The Derg’s identity project was destructive rather than constructive. The Marxist-Leninist ideology the Derg adopted after 1976 posited a socialist Ethiopian identity organised around class rather than ethnicity, and framed the imperial-Christian-Amharic identity as feudal-reactionary. The land reform, the literacy campaign, and the mass mobilisation apparatus did produce a version of pan-Ethiopian identity for the populations that participated in them — the peasants who received land, the students who taught literacy, the militia members who served in the war with Eritrea and Somalia — but this identity was thin and did not survive the Derg’s collapse in 1991. The identity project did not build a durable pan-Ethiopian nation; it hollowed out the imperial one and left, at the point of the EPRDF’s arrival in Addis Ababa, a population without an operational shared identity.
The EPRDF’s identity project was, in operational terms, the codification of separate identities into constitutional form. The 1995 Constitution’s identification of the “Nations, Nationalities and Peoples of Ethiopia” as the constituent subjects (Article 8), the federal architecture organised around nationality-based regional states, and the right of self-determination up to secession (Article 39) all rested on the theoretical proposition that Ethiopian identity was, in the EPRDF’s phrase, “unity in diversity” — that the country was made up of multiple national identities that would coexist within the federal state. The proposition has been the subject of the previous series in this sequence (“The Hollow Pillar”), and this series will not re-examine it in constitutional detail. What matters here is that the EPRDF framework did not construct a pan-Ethiopian identity at the national level; it institutionalised the sub-national identities as constitutionally sovereign, and left the pan-Ethiopian identity as an aspirational rather than politically operative fact.
The Prosperity Party’s Medemer project since 2018 has been the first attempt at pan-Ethiopian identity construction under a post-1991 regime. Abiy Ahmed’s rhetoric has explicitly invoked the language of national unity, shared purpose, and pan-Ethiopian community. The project has produced symbolic outputs (the Green Legacy reforestation, the Sheger Park development, the Dine for Generations programme) and administrative outputs (the merger of the EPRDF constituent parties into the unitary Prosperity Party, the Amhara Special Forces dissolution, the various centralisation moves). What it has not produced is a shared pan-Ethiopian identity that operationally holds the population together. The evidence: three simultaneous armed conflicts (Amhara, Tigray, Oromia) within the state Abiy governs, each involving substantial populations that do not accept the state’s authority as their own. The Medemer project has been, in Article 23’s framing, right instinct badly executed. The instinct — that Ethiopia needs a pan-Ethiopian identity — is correct. The execution has not produced the identity.
The Eritrean case
The 1993 Eritrean separation is the most dramatic single case of Ethiopian identity failure. What made the separation possible was the failure of the Ethiopian imperial and Derg-era identity project to construct a version of Ethiopian-ness that Eritreans could accept as their own.
Eritrea had been part of the Ethiopian imperial territory only intermittently before 1890, and the fifty years of Italian colonial administration (1890–1941) had produced a distinct Eritrean political culture. The 1952 federation, imposed by UN resolution against significant Eritrean nationalist preference for independence, was a nominal recognition of Eritrean distinctiveness within an Ethiopian framework. The 1962 dissolution of the federation, imposed by Haile Selassie’s government through coercion of the Eritrean assembly, was the moment at which the imperial framework demonstrated that it would not accommodate Eritrean identity within its own structure. The thirty-year war that followed was the operational consequence of that decision. By 1991, the war had been won by the Eritrean People’s Liberation Front, and the 1993 referendum ratified an outcome that Ethiopian identity policy had made inevitable.
The counterfactual is what a proper Ethiopian identity project would have looked like from 1952. The imperial state could have accommodated Eritrean distinctiveness within a federal framework that recognised Eritrean political culture, Tigrinya and Arabic as official languages of Eritrean administration, the Eritrean flag as a valid regional symbol under the Ethiopian imperial standard, and the Eritrean assembly as a legitimate representative body coordinating with the imperial parliament. This is roughly what post-war Japan did with Okinawa (with problems that continued into the 2020s but without producing separation), what post-1949 China did with Xinjiang and Tibet (with more serious problems but again without producing separation), and what Malaysia does with Sabah and Sarawak (with continuing tensions but again without separation). None of these are unblemished successes. All of them are cases in which the central state made active identity-construction and accommodation efforts that Ethiopia did not.
Instead, the Ethiopian imperial state pursued assimilationist policies: Amharic replaced Tigrinya and Arabic in Eritrean administration through the 1950s; the Eritrean flag was abolished in 1959; the federation was dissolved in 1962; the war of independence began in 1961 and continued until 1991. Each step deepened Eritrean rejection of the imperial identity project rather than incorporating Eritrea into it. The Derg’s period intensified the war and made no serious identity construction effort. The 1993 referendum ratified the failure.
The parallel with Korea is instructive. North and South Korea were divided in 1945 by external imposition (Soviet and American occupation zones), and the division has been sustained by external great-power politics as well as internal regime differences. Ethiopia and Eritrea were divided by domestic policy failure sustained over forty years. The Korean case is not one that either Korean society regards as fully legitimate; both maintain formal commitments to eventual reunification (variously credible). The Ethiopian case is one that the Eritrean state regards as fully legitimate (with corresponding formal recognition and diplomatic relations, when those have not been suspended by conflict) and the Ethiopian state regards, in operational terms, as an increasingly regretted historical event whose reversal would be politically explosive.
The salt-and-neighbor image Zef used in developing this series is exactly right. Populations that are politically separated but geographically adjacent, culturally overlapping, and family-linked can, over generations, drift into treating each other as foreign — or can, through active state policy on both sides, maintain the shared political-community sense that Ethiopia and Eritrea failed to maintain. What the Asian comparators show is that such policy is possible and that it works when sustained. What Ethiopia’s record shows is what happens when the policy is not made.
What comes next
Article 3 takes up the Asian developmental model as a set of common features that the comparators shared. Article 4 argues why identity construction, not restoration, is the correct frame — the point Zef and I worked through in the initial planning conversation and that is central to the series. Part II then begins the case-by-case examination with Meiji and post-war Japan.
The identity question is what this series exists to examine. The subsequent articles will develop the argument through the specific cases. The general proposition established here — that identity is a political fact constructed by state action, that its absence forecloses the political-economic capacities of the modern state, and that Ethiopia has not constructed it — is the foundation on which the rest of the series rests.