The Missed Opportunity: Costs Paid, Benefits Never Received

Thesis

Between 1961 and 1993, three East Asian countries transformed themselves from war-devastated, resource-poor, mostly peasant societies into industrialised producers integrated into the global economy. South Korea’s per-capita GDP rose from approximately $72 in 1961 to over $8,000 by 1993.¹ Taiwan’s rose from a comparable base to over $10,000. Singapore’s rose from $500 in 1965 (the year of independence from Malaysia) to over $18,000 by 1993. Japan had made a similar transition twice, once between 1868 and 1912, and again between 1945 and 1975 under American occupation and its aftermath. China, starting later and starting from a much larger and poorer base, began its own version of the same transformation in 1978 under Deng Xiaoping, and by 2020 had become the world’s second-largest economy. These transformations were not accidents. They were the product of specific policy choices, sustained over decades, executed by state institutions organised for the purpose.

The transformations came at substantial cost. The Park Chung-hee period in South Korea (1961–1979) was authoritarian, produced the Yushin constitution’s suspension of political freedoms, imprisoned thousands of opposition figures, and ended with Park’s assassination by his own intelligence chief.² Chiang Ching-kuo’s Taiwan operated under martial law from 1949 to 1987, with the White Terror against political opposition producing an estimated 3,000 to 4,000 executions and tens of thousands of imprisonments.³ Deng’s China produced the Tiananmen Square killings of June 1989. Lee Kuan Yew’s Singapore kept opposition leaders in Internal Security Act detention for decades and restricted the press through the 1980s and 1990s.⁴ The Japanese Meiji Restoration was preceded by the Boshin War, followed by the Satsuma Rebellion, and produced a state that would, by the 1930s, become militarist and imperial.

The Asian transformations were, in short, not costless. They were paid for in political prisoners, suppressed opposition, forced integration, delayed democratisation, and in the Japanese case eventually in war. What made them transformations rather than merely repressions is what the price bought. Each of the Asian states delivered, over the course of one generation or two, the industrial base and the working nation that the price was supposed to purchase. The populations paying the price got, in return, per-capita incomes ten to a hundred times higher than they had started with, and — critically — pan-national identities under which their descendants could be Korean, Taiwanese, Singaporean, Japanese, or Chinese in a way that was politically operative in the modern world.

Ethiopia is the counter-case this series exists to examine. Between 1974 and 2026, Ethiopia paid substantial versions of all the prices the Asian comparators paid. The Derg’s Red Terror killed between 30,000 and 150,000 Ethiopians for political reasons; the war with the EPLF and TPLF killed hundreds of thousands more; the 1984–85 famine killed an estimated 400,000 to 1 million. The EPRDF period from 1991 saw the 2005 election crackdown, the post-2005 institutionalised suppression of opposition under the Anti-Terrorism Proclamation, and the resumption of famine as counter-insurgency instrument. The post-2020 period has seen the Tigray war (an estimated 600,000 to 800,000 dead), the Amhara conflict (ongoing as of mid-2026), and the OLA insurgency in Oromia. Ethiopian per-capita income in 2026 remains below $1,500, having risen from approximately $200 in 1961.⁵ The Asian comparators, starting from a similar or worse position, are now between five and thirty times wealthier per capita. The Ethiopian population has paid the price. The Ethiopian population has not received what the price was supposed to buy.

The bargain and its terms

The transactional logic of twentieth-century developmentalism was straightforward, if brutal. In exchange for concentrated executive authority sufficient to override sectional resistance, the state would deliver rapid industrialisation, universal education, expanded urban employment, and — most importantly — the pan-national identity that would make the population a nation rather than a collection of tribes, regions, and religions. The bargain was not democratic in its terms. Populations were not asked to consent to it in most cases; they were required to submit to it. But the bargain had an implicit contract: the executive would use its authority for the productive work of nation-building, and the population would tolerate the authority for the duration of the work.

Every one of the Asian comparators kept this bargain. Meiji Japan by 1912 had produced a nation whose subjects understood themselves as Japanese, whose economy could compete with European powers, and whose military had defeated both China (1895) and Russia (1905). Post-war Japan by 1975 had produced the second-largest economy in the world, a democratic constitutional order under the 1947 constitution, and a national identity that has since survived multiple political-economic shocks intact.

Park’s Korea by 1979, when Park was killed, had produced the industrial base that would carry South Korea through the 1980s democratisation transition and into the 1990s membership of the OECD. Chiang Ching-kuo’s Taiwan by 1988, when Chiang died, had produced the semiconductor and electronics industries that continue to make Taiwan globally strategic, and had begun the democratisation process his son and successor would complete. Deng’s China by his own death in 1997 had produced the export-manufacturing base that would, over the next two decades, lift several hundred million Chinese out of poverty and produce the world’s second-largest economy.

Lee Kuan Yew’s Singapore by his retirement as Prime Minister in 1990 had produced a per-capita income comparable to Britain’s, a functioning multi-ethnic society organised around a constructed Singaporean identity, and a state whose institutions would survive his eventual death in 2015 without significant discontinuity.

Ethiopia’s bargain, by contrast, has failed at every regime change and at every internal transition. The imperial state produced modest infrastructural gains between 1930 and 1974 but did not industrialise, did not universalise education, did not construct pan-national identity, and did not survive the 1974 revolution. The Derg’s fifteen years of authoritarian developmentalism produced land reform (a real achievement), universal literacy campaigns (partial), and Red Terror (unambiguous), but not industrialisation and not durable national institutions. The EPRDF’s twenty-seven years produced GDP growth that peaked at over 10 percent annually, expansion of higher education from 18,000 students to over a million, and infrastructural investment on an unprecedented scale, but not durable industrial base, not political inclusion, and not the pan-national identity that would have carried the growth beyond the specific political configuration that produced it. The Abiy government’s eight years to date have produced constitutional consolidation, symbolic re-founding through the Medemer framework, and macroeconomic stabilisation, but simultaneous civil wars in three regional states and further deterioration of pan-national identity.

Ethiopia has paid the developmental authoritarianism’s price. Ethiopia has not received what the price was supposed to buy. This is what the title of this series means.

The three failures

The gap between Ethiopian cost and Ethiopian benefit is not a single failure but three interlocking failures, each traced in one of the three series this article opens.

The monetary failure — analysed in “The Birr and the Pretence of Knowledge” — is the recurrent monetary disorder that has produced, across three regimes, five currency reforms, chronic inflation, capital controls, and the failure to develop the domestic banking and financial infrastructure that industrial economies require.

The constitutional failure — analysed in “The Hollow Pillar” — is the recurrent constitutional pathology that has produced, across three regimes and five constitutional instruments, executive sovereignty that is not bound by the constitutional text.

The identity failure — the subject of this series — is the failure to construct a pan-Ethiopian national identity capable of holding a multi-ethnic, multi-religious population together as one political community with shared purposes. The imperial identity Ethiopia inherited was too narrow. The Derg’s revolutionary identity was too destructive. The EPRDF’s ethnic-federal identity was too centrifugal. The Prosperity Party’s Medemer identity has been too symbolic.

Each of these failures is a component of the larger story. Each has its own explanatory logic and its own institutional expression. But they share a single underlying cause: the political-economy configuration in which unconstrained executive authority has been more valuable to successive Ethiopian regimes than the constructive nation-building work that would have produced a functioning state. The Asian comparators had similarly configured political economies — authoritarian, executive-centered, coalition-narrow — but their executives used the authority to construct rather than to consolidate. Ethiopian executives, across three regimes, have used the authority to consolidate against their own populations rather than to construct with them.

Why this comparison, and why now

Two objections to the comparison have to be addressed before the series proceeds.

The first objection is that Ethiopia is not comparable to the Asian cases because Ethiopia has specific features (multi-ethnic composition, land-locked status since 1993, agrarian economic base, low education inheritance) that the Asian states did not share. The reply is partial. Ethiopia is not identical to any of the Asian cases; no country is identical to any other. But Ethiopia does share, in various combinations, features with several of them. Multi-ethnicity: Singapore and Malaysia in acute form; Indonesia in complex form. Land-locked status: not shared with any Asian case, and a genuinely distinctive feature of the Ethiopian condition (and one the 1993 Eritrean separation made worse, discussed in Article 20). Agrarian economic base at the start of the transformation: shared by all the Asian comparators except Singapore. Low education inheritance: shared by Korea, Taiwan, and China at their respective starting points. The comparison is not exact; it is instructive.

The second objection is that the Asian model was made possible by specific Cold War conditions (American security guarantees, American market access, capital transfers) that Ethiopia did not and cannot enjoy in the 2020s. This objection has more weight. American security guarantees and market access did contribute to the Asian transformations; the specific configuration of the Cold War produced opportunities that later developing countries have not had. But the Asian model was not solely produced by American assistance. Land reform, education investment, meritocratic bureaucracy, export orientation, and identity construction were domestic decisions taken by domestic governments. Ethiopia has received substantial external assistance since 1991, including major infrastructural investment from China, budget support from the EU and multilateral institutions, and humanitarian support from a wide range of donors. The volume of assistance is comparable, in relative terms, to what the Asian comparators received. What has been different is what the assistance has been used for. Ethiopia has not used its external resources to build the institutions the Asian comparators built. That is the failure this series will examine.

What the series argues

The series argues, across its twenty-eight articles, four propositions.

First, that the Asian developmental model is real, is well-understood, and has been implemented in conditions comparable to Ethiopia’s. The Asian cases are not exotic. The technical and institutional literature on them is extensive. Ethiopian policymakers and intellectuals have known about them since at least the 1990s, when Meles Zenawi published his own extensive engagement with the East Asian developmental-state literature. The model is not secret.

Second, that the Asian model came at costs Ethiopia is already paying. Political prisoners, suppressed opposition, restricted press, forced resettlement, state violence, delayed democratisation — Ethiopia has paid all of these under the Derg, all of these under the EPRDF, and continues to pay several of them under the Prosperity Party. The proposition that Ethiopia cannot afford the Asian model because of its costs is inverted: Ethiopia is already paying the costs, and the argument is not whether to pay them but whether to receive anything in return.

Third, that the central missing element in Ethiopia’s developmental attempts has been identity construction. Every Asian comparator built pan-national identity out of the raw materials available. Meiji Japan constructed Japanese identity from imperial cult, State Shinto, universal education, and universal conscription. Post-war Japan reconstructed it from constitutional democracy, economic recovery, and pacifism. Korea built Korean identity from land reform, education, and industrial modernity. Singapore built Singaporean identity from housing policy, bilingual education, national service, and public culture. China built Chinese identity from Party discipline, universal education, and economic transformation. Ethiopia has not attempted anything comparable at scale. This is the central missing element, and Article 4 of this series will examine why.

Fourth, that the constructing of Ethiopian identity remains possible and must be undertaken. The construction cannot be restoration; the imperial identity Ethiopia had before 1974 was not consented to by the populations it tried to hold, and cannot be. The construction cannot be ethnic-federal division; the 1995 architecture has produced the centrifugal pressure the country is presently absorbing. The construction has to be new, drawing on all Ethiopian inheritances, sustained across generations by state action of a kind Ethiopia’s regimes have not yet undertaken. What that construction would look like is the subject of the final part of this series (Articles 25–28).

What the series does not argue

The series is not an apology for authoritarianism. Ethiopia is not going to industrialise or construct identity or transform itself under the current configuration of executive power; the previous two series have argued at length that the executive-centered political economy is precisely the problem. What the series argues is that the Asian comparators combined executive authority with productive use of that authority, and that Ethiopia’s regimes have combined executive authority with unproductive use of it. The remedy is not more executive authority. The remedy is redirected executive authority.

The series is not an argument that Ethiopia should copy any specific comparator. Singapore’s HDB Ethnic Integration Policy is instructive, not directly transferable; Park’s chaebol model is instructive, not directly transferable; Meiji’s State Shinto is instructive, not transferable at all. What the comparators offer is a set of successful examples of the kind of institutional and identity work Ethiopia has not done, from which specific lessons can be extracted and adapted.

The series is not an argument that Ethiopia’s problems can be solved by policy alone. The political-economy configuration that has produced the failures is deep and self-reinforcing. Changing it requires not just better policy but different politics. The identification of what needs to be done is easier than the identification of who will do it and under what political configuration. This series will offer both — the what and the possible who — but the who remains harder.

The Eritrea case as illustration

One case will thread through the series as illustration of the identity failure at its most consequential: the Eritrean separation. The 1993 referendum produced Ethiopian recognition of Eritrean independence. The recognition was procedurally correct; Eritrean self-determination against imperial and Derg-era rule was legitimate; the referendum was conducted under UN supervision and produced 99.8 percent affirmative vote.

The referendum was also the operational outcome of forty years of Ethiopian failure to construct an identity that could hold Eritrea inside the country. Under the imperial framework, Eritrea was administered as a fourteenth province with progressive assimilationist policies (dissolution of the federation 1962, banning of Tigrinya and Arabic in Eritrean administration 1956–1959, imposition of Amharic curriculum) that the Eritrean population did not accept. Under the Derg, the war of independence intensified and became the central strategic problem of the regime for eighteen years. By 1991, when the EPLF took Asmara and the EPRDF took Addis Ababa, the political fact of Eritrean separation was already established; the referendum ratified it.

The Asian comparators would have handled this differently. Meiji Japan integrated Hokkaido, Okinawa, Taiwan, and later Korea through a combination of administrative modernisation, education standardisation, and identity construction (variously successful, and in the case of Korea unsuccessful in ways that produced the eventual 1945 separation, but a serious attempt). Post-war Japan integrated the outer prefectures with democratic institutions and shared national identity built on pacifism and economic recovery. China integrated Tibet, Xinjiang, and Inner Mongolia through a combination of coercion, administrative modernisation, and economic incorporation — problematically, in ways this series will not endorse, but effectively. Singapore integrated a Chinese-majority population with substantial Malay and Indian minorities through the identity construction this series will examine in detail. Malaysia held together a comparable multi-ethnic composition through the New Economic Policy and Bumiputera framework, imperfectly but durably.

Ethiopia’s failure with Eritrea was not the acceptance of the referendum. It was the forty years of policy that made the referendum inevitable. The imperial state under Haile Selassie did not construct an identity Eritreans could accept. The Derg did not construct one either. The EPRDF, when it came to power in 1991, accepted the separation as a completed fact rather than attempting to construct a shared future political community. This is what the identity failure looks like in operational terms. This is the price Ethiopia paid, and continues to pay in 2026, without receiving what the price was supposed to buy.

The Korean parallel is instructive but not exact. North and South Korea were divided in 1945 by external imposition, and the division has been sustained by external and internal factors that neither Korean society regards as fully legitimate. Ethiopia and Eritrea are divided by domestic decision, and the division is regarded as legitimate by one party (Eritrea) and increasingly regretted by the other (Ethiopia). The parallel is that in both cases, populations with substantial shared history, religion, language, and family are politically separated in ways that reduce both parties’ welfare. The difference is that the Ethiopian side of the separation was chosen by Ethiopian decision.

Article 20 will examine the Eritrean question in detail. For the present article, the case establishes the pattern that the series will trace: identity failure produces separation, separation produces cost, cost is paid, and the return is not received.

What comes next

The next three articles in Part I establish the analytical framework: what identity means as a political fact (Article 2), what the Asian developmental model is in its common features (Article 3), and why construction rather than restoration is the correct frame (Article 4). Part II takes up Japan as the original case. Part III examines the Cold War Tigers. Part IV takes up Singapore as the central multi-ethnic case. Part V examines multi-ethnic variants and cautionary failures. Part VI applies the framework to Ethiopia’s specific historical turns. Part VII synthesises and prescribes.

The argument is long. The failure it examines has taken ninety-five years to produce, and the answer to it will take generations to construct. This series is the map of the territory, not the journey through it.

---

References

Footnotes

1. On Korean per-capita GDP in 1961 and after, see Park Chung-hee era literature; Uk Heo et al., “The Political Economy of South Korea: Economic Growth, Democratization, and Financial Crisis,” Maryland Series in Contemporary Asian Studies 193, no. 2 (2008); Miracle on the Han River literature. ↩

2. On the Yushin constitution and Park’s assassination, see the Park Chung Hee Era literature; Ezra F. Vogel et al., eds., The Park Chung Hee Era: The Transformation of South Korea (Harvard, 2011). ↩

3. On the White Terror in Taiwan, see Denny Roy, Taiwan: A Political History (Cornell, 2003); Taiwan Transitional Justice Commission reports 2018–2022. ↩

4. On Lee Kuan Yew’s Internal Security Act detentions, see Michael D. Barr, The Ruling Elite of Singapore (I.B. Tauris, 2014); Cherian George, Singapore, Incomplete (Woodsville News, 2017). ↩

5. On Ethiopian GDP per capita, see World Bank and IMF Ethiopia data 1961–2025; recent CSIS analysis of Ethiopia’s 2026 election trajectory. ↩