Meles's Constitutional Theory and Practice: Revolutionary Democracy in Its Own Words

Thesis

Meles Zenawi was Prime Minister of Ethiopia from 1995 until his death in office on 20 August 2012. He was also, more importantly for the purposes of this series, the principal intellectual architect of the post-1991 Ethiopian state. His writings, speeches, and policy interventions over twenty-one years constitute the most fully articulated theory of what the 1995 Constitution was for, what kind of state Ethiopia should be, and what role constitutional structure should play in producing developmental outcomes. The theory has a name in Meles’s own vocabulary: revolutionary democracy.¹ It is an internally coherent intellectual framework that, on its own terms, justifies most of what this series has identified as the operational pathologies of the 1995 settlement.

The article’s argument — and this is the steelman that the series owes to Meles as the most serious defender of the constitutional order that produced it — is that Meles’s revolutionary democracy is not a hypocritical cover for authoritarianism but a substantively distinct constitutional theory. It rejects the Madisonian framework not because it does not understand it but because it disagrees with it. It rejects the Hayekian framework not because it has not read it but because it has read it and decided that the rule-of-law-as-predictable-general-rules cannot, in the conditions of Ethiopia in 1991, produce the developmental outcomes that the country needs. It rejects the Acemoglu-Robinson framework not because it does not see the appeal of inclusive institutions but because it believes that, in conditions like Ethiopia’s, the developmental state must produce growth before institutional inclusion can be afforded. The theory is not stupid. It has produced real intellectual followers, including the constitutional scholars who continue to defend the 1995 settlement after Meles’s death. The series engages it as the strongest version of the case for the constitutional order this series has critiqued.

The intellectual origins

Meles Zenawi was born in 1955 in Adwa, in Tigray; left medical school at Haile Selassie I University in 1974 to join the TPLF’s armed struggle; held leadership positions in the TPLF from the late 1970s; became the TPLF’s chair in 1989; led the EPRDF coalition to victory in 1991; served as President of the Transitional Government from 1991 to 1995; and was Prime Minister of Ethiopia from 1995 to 2012.² He published academic work throughout his political career, including the unfinished 2006 paper “African Development: Dead Ends and New Beginnings,” delivered as a lecture at Columbia University, and various policy papers under the EPRDF’s intellectual umbrella.³

The intellectual sources of Meles’s theory were several. Marxist-Leninist theory, particularly Lenin’s writings on the national question and on the dictatorship of the proletariat, provided the foundation; the TPLF had been founded as a Marxist-Leninist organisation in 1975 and Meles’s early intellectual formation was within this framework.⁴ The developmental-state theory associated with East Asian growth experiences (Japan, South Korea, Taiwan, Singapore) provided the model after the Cold War; Meles became a serious student of East Asian development and explicitly modelled aspects of Ethiopian policy on East Asian precedents.⁵ The critique of neoliberal economics, including the Washington Consensus reforms that Meles believed had failed in Africa, provided the framework for resisting IMF and World Bank conditionalities.⁶ And the African nationalist tradition, including Nyerere’s ujamaa socialism and various other post-colonial state-building projects, provided the political ancestry.

The synthesis Meles produced from these sources he called revolutionary democracy. The fullest statements of the theory appear in the EPRDF’s internal documents and in Meles’s published academic work; partial popular versions appear in his many parliamentary speeches and press conferences. The theory has not been the subject of a single canonical statement, but its main elements can be reconstructed from his various writings.⁷

The theory in summary

Revolutionary democracy holds:

First, that liberal-democratic constitutionalism in Africa has, with rare exceptions, produced extractive rather than developmental outcomes. Multiparty competition has produced ethnic political competition that destabilises the state. Independent courts have, in conditions of weak rule of law, become sites for elite manipulation rather than constraints on it. Free media have, in conditions of low literacy and weak civic culture, amplified destabilising messages rather than producing informed citizenship. The Western constitutional template, applied in African conditions, has produced not the inclusive institutions Acemoglu and Robinson celebrate but the extractive politics they decry.⁸

Second, that the conditions for inclusive institutions cannot be created by constitutional design alone. Inclusive institutions require a social base — a literate, prosperous, organised citizenry with the means to demand institutional inclusion. African states emerging from colonialism, civil war, or imperial rule lack this base. Drafting constitutional documents that assume the base exists is a category error. The state must produce the social base first; the constitutional inclusion can follow.

Third, that the developmental state, organised under a vanguard party with strong central direction, is the political form that produces the social base. East Asian growth (1960s–1990s) was produced by such states. African growth, where it has occurred, has been produced by similar arrangements (Rwanda after 1994, Ethiopia from 1991). Multiparty liberal-democratic states have not produced comparable growth in Africa. The developmental-state model is empirically vindicated by comparison.

Fourth, that the developmental state requires constitutional and institutional features that look authoritarian by Western standards but are necessary for the developmental project. These include strong executive authority, restricted electoral competition (in the sense of preventing destabilising opposition organisation, not in the sense of preventing all opposition), state ownership of strategic resources (including land, Article 40), and a vanguard party with disciplined cadres operating across state institutions.⁹

Fifth, that constitutional rights and democratic forms are not abandoned in this framework but are sequenced. The state delivers economic growth and social development first; democratic deepening can occur over time as the social base for it develops. This is not a permanent suspension of constitutional democracy; it is a strategic ordering.¹⁰

The theory is internally coherent. It is empirically anchored to East Asian and Rwandan comparators. It directly addresses the criticisms that this series has been making and offers a substantive response to each.

Application to the 1995 Constitution

Meles applied revolutionary democracy to the 1995 Constitution in ways that explain features of the constitutional order this series has identified as problematic.

Why the House of Federation rather than a constitutional court. Constitutional review by a political body, in Meles’s framework, is preferable to review by professional judges because the constitution’s interpretation is a political question that should be answered by political representatives, and because professional judges in conditions of weak rule of law are more vulnerable to elite manipulation than political representatives elected by ethnic constituencies. The Madisonian preference for judicial review reflects the institutional conditions of mature constitutional democracies and does not transfer to post-civil-war Ethiopia.¹¹

Why Article 39 (secession). The constitutional right of self-determination, in Meles’s framework, was the price of peaceful federation. Without it, the various nationality movements would not have agreed to the federation; the federation would have collapsed into renewed civil war. The right is constitutional, but the conditions for its peaceful invocation are political — and the political conditions are largely the federation’s developmental performance. A federation that delivers growth and development to its constituent nationalities will not face secession invocation; a federation that fails to deliver will. The right is, in this framework, an incentive structure rather than a permanent threat.¹²

Why Article 40 (land). State ownership of land was not an ideological holdover but a developmental necessity. Private land ownership would, in the conditions of Ethiopian agriculture, produce rapid landlessness and rural impoverishment that would destabilise the developmental project. State ownership permits the developmental state to direct large-scale agricultural investment, to manage rural-urban transition, and to prevent the social differentiation that would produce demands for political change before the social base for inclusive institutions had developed. The constitutional commitment to state land ownership is, in this framework, a developmental-state instrument.¹³

Why the 2005 crackdown. The post-2005 prosecutions, in Meles’s framework, were not violations of the constitutional order but defences of it. The CUD had organised an attempt to overthrow the constitutional government through extra-electoral means; the constitutional response — prosecution under existing law, imposition of order, restoration of normal political functioning — was justified. The opposition’s electoral gains had been recognised in the 109 seats they won; the post-election protests had crossed into extra-constitutional territory; the response was proportional.¹⁴

Why the 2009 Anti-Terrorism Proclamation. The legal infrastructure for counter-terrorism, in Meles’s framework, was necessary both for the actual terrorism threats Ethiopia faced (Al-Shabaab, regional al-Qaeda) and for the developmental-state requirement of political order. Opposition political organisation that crossed the line into armed activity (OLF, ONLF, Ginbot 7) had to be addressed legally; the ATP provided the framework. The application to journalists who reported on these organisations was, in Meles’s framework, sometimes overbroad but not in principle constitutionally problematic.¹⁵

The theory provides coherent justification for each of these features. It is not a defence of every operational choice — Meles himself acknowledged, in his 2010 EPRDF congress speech, that the 2005 crisis had been mishandled in specific respects.¹⁶ But the theory provides the framework within which the operational choices were made, and it provides intellectual resources for defending most of them.

The reply

The reply to revolutionary democracy operates on three levels.

First, empirical. The empirical case for the developmental-state model in Ethiopia is mixed. Ethiopia’s GDP growth from 2003 to 2018 was indeed exceptional by African standards, averaging over 10 percent in some years.¹⁷ But the growth was concentrated in particular sectors (construction, services, some manufacturing) and did not produce the broad-based social transformation the theory promised. The Ethiopian middle class that emerged in the 2010s remained small relative to the population; rural poverty remained pervasive; the agricultural transformation the theory predicted did not occur at the scale projected. By 2026, hyperinflation, abject poverty at 43 percent, and multidimensional poverty above 70 percent indicate that the developmental-state delivery has been incomplete.¹⁸ The empirical record does not refute the theory’s claim that developmental-state policies can produce growth, but it complicates the claim that they did produce the social base for inclusive institutions in the timeframe Meles projected.

Second, comparative. The East Asian comparators — Japan, South Korea, Taiwan, Singapore — produced developmental growth under authoritarian or semi-authoritarian conditions and transitioned to constitutional democracy after the growth had occurred. The model Meles invoked is, on its own terms, a model that ends in democratisation. South Korea democratised in 1987 after twenty-five years of developmental-state growth; Taiwan democratised in 1996; Singapore remains partial. Meles’s framework called for democratisation to follow growth. The transition never came under his successor Hailemariam Desalegn (2012–2018), and the 2018 transition under Abiy was produced by protest mobilisation and elite manoeuvring rather than by the planned democratic deepening the theory predicted. The model, applied to Ethiopia, did not produce its own predicted endpoint.

Third, conceptual. Revolutionary democracy’s framework assumes that the political organisation conducting the developmental project — the EPRDF as vanguard party — is institutionally capable of self-correction when the developmental project falters, and that the developmental project will not become an end in itself rather than a means to inclusive institutions. The Ethiopian record suggests both assumptions were optimistic. The EPRDF did not self-correct effectively in the 2010s; the post-2014 protest mobilisations that produced the 2018 transition were not absorbed into a planned democratisation but precipitated a leadership crisis. And the developmental project, once underway, became the framework’s justification rather than its instrument: the framework was no longer “developmental state until inclusive institutions are possible” but “developmental state because inclusive institutions threaten the developmental state.”

Meles after Meles

Meles died in August 2012 of an undisclosed illness, in office.¹⁹ His successor, Hailemariam Desalegn, governed from 2012 to 2018 under the framework Meles had established but without Meles’s intellectual authority within the EPRDF. The framework persisted institutionally but lost its intellectual coherence. The post-Meles EPRDF was a coalition that no longer had a unifying ideological vision; the developmental-state project continued but the political organisation lacked the discipline that Meles had imposed.

The 2014–2018 protest mobilisations (subject of the next article) were partly a response to the post-Meles drift. The grievances that fuelled the protests — land dispossession, urban unemployment, political restriction, perceived ethnic favouritism — were grievances that the developmental state had not addressed and that the framework had not predicted. The 2018 transition under Abiy was, in operational terms, a partial repudiation of the Meles framework: the move from a coalition party (EPRDF) to a unitary one (Prosperity Party), the rhetoric of pan-Ethiopian unity over ethnic federalism, the partial economic liberalisation, the rapprochement with Eritrea — each of these was, in Meles’s framework, a deviation from the developmental-state model.

Whether the Abiy framework is more or less successful than Meles’s by the end of the 2020s is a question this series will not resolve. What is clear by 2026 is that the post-Meles experiment has produced significant new instability (the Tigray war, the Amhara/Oromia/Sidama crises) while the developmental indicators have moved in mixed directions. The Meles framework’s defenders — Solomon Negussie, Andreas Eshete, and others associated with the EPRDF intellectual tradition — argue that the post-2018 problems vindicate the Meles framework by demonstrating what happens when its discipline is abandoned. The Abiy government’s defenders argue that the Meles framework had already failed and that the post-2018 deconstruction was overdue.

This series takes a different view: that both Meles’s framework and Abiy’s framework operate within the constitutional architecture that the 1995 settlement established, that the architecture has structural pathologies that neither framework has addressed, and that the choice between revolutionary democracy and pan-Ethiopian unity is, at the constitutional level, less important than the architecture both frameworks share.

Steelman of revolutionary democracy

The strongest version of the revolutionary-democracy defence is the one this article has been trying to present. The theory holds that the constitutional design and operational pattern this series criticises were not failures of constitutionalism but features of a different constitutional vision: developmental-state constitutionalism, sequenced toward eventual democratic deepening, organised around vanguard-party direction, accepting of restricted political competition for the duration of the developmental phase. The theory has serious advocates, real empirical referents in East Asia, and substantive intellectual content. It cannot be dismissed as cynicism dressed in theoretical vocabulary.

The reply this series has been making — that the constitutional architecture has produced predictable pathologies regardless of which framework occupies it, that the developmental delivery has not vindicated the theory’s promises, and that the sequencing toward democratisation has not occurred on the timeline the theory projected — is a substantive disagreement, not a refutation. The disagreement is at the level of constitutional theory and political-economy prediction, not at the level of empirical fact alone. Reasonable observers can disagree about which framework better captures the Ethiopian condition. This series stands on its disagreement while acknowledging the force of the position it disagrees with.

Multi-causal accounting

Meles’s framework was not the only intellectual influence on the 1995 settlement; Andreas Eshete’s more philosophical defence of ethnic federalism, Fasil Nahum’s institutional analysis, and the contributions of other Constitutional Commission members all played a role. The Meles framework was not the only intellectual force on post-2005 securitisation; the post-9/11 international context, the immediate political pressures from the 2005 crisis, and the operational requirements of the developmental project all contributed. The framework is one explanatory factor; the broader institutional and political-economy conditions are others.

What the article establishes

Meles’s revolutionary democracy is the strongest available theoretical defence of the 1995 Ethiopian constitutional order. It is internally coherent, empirically referent, and substantively addressed to the criticisms this series has been making. Its empirical track record is mixed; its predicted endpoint (sequenced democratisation) has not been achieved; its post-Meles institutional continuation has been incomplete. The framework operates within the constitutional architecture this series has criticised and shares the architecture’s structural features. Engaging the framework at its strongest is what the series owes to Meles as the constitutional order’s principal author. Disagreeing with the framework on substantive grounds is what the series concludes after the engagement.

The next article takes up the protest mobilisations of 2015–2018 that produced the most significant test of the post-Meles constitutional order and led to the 2018 transition.

---

References

Footnotes

1. On revolutionary democracy as a concept, see EPRDF internal documents available in translation; René Lefort, “The Theory and Practice of Meles Zenawi: A Response to Alex de Waal,” African Affairs 112, no. 448 (2013); Alex de Waal, “The Theory and Practice of Meles Zenawi,” African Affairs 112, no. 446 (2013). ↩

2. Biographical material from various sources; see Paulos Milkias, Haile Selassie, Western Education, and Political Revolution in Ethiopia (Cambria Press, 2006); Aregawi Berhe, A Political History of the Tigray People’s Liberation Front. ↩

3. Meles Zenawi, “African Development: Dead Ends and New Beginnings,” lecture at Columbia University, 2006 (unpublished but circulated). ↩

4. On TPLF Marxism-Leninism, see John Young, Peasant Revolution in Ethiopia, ch. 5; Aregawi Berhe, A Political History. ↩

5. On Meles’s East Asian framework, see Alex de Waal, The Real Politics of the Horn of Africa (Cambridge: Polity, 2015), and his Foreign Affairs articles on Meles. ↩

6. Meles, “African Development”; Joseph Stiglitz’s writings on Meles after their professional interactions. ↩

7. On the construction of the theory, see de Waal, “The Theory and Practice of Meles Zenawi.” ↩

8. This is the most controversial claim in revolutionary democracy and one of its most explicit; see EPRDF internal documents. ↩

9. On the vanguard-party requirement, see EPRDF programmatic documents; Sarah Vaughan and Mesfin Gebremichael, Rethinking Business and Politics in Ethiopia (2011). ↩

10. The sequencing argument is most explicit in Meles’s parliamentary speeches and in the EPRDF’s later programmatic documents. ↩

11. On the design choice for the House of Federation, see Fasil Nahum and Andreas Eshete writings. ↩

12. On Article 39 as incentive structure, see Andreas Eshete’s writings; Aregawi Berhe. ↩

13. On Article 40 as developmental-state instrument, see Tom Lavers’ work; the official EPRDF policy documents. ↩

14. For Meles’s framing of the 2005 crisis, see his parliamentary speeches 2005–2006. ↩

15. For Meles’s framing of the 2009 ATP, see his parliamentary speech introducing the legislation. ↩

16. Meles’s 2010 EPRDF congress speech. ↩

17. On Ethiopian growth statistics, see Christopher Adam and Stefan Dercon work; IMF Article IV consultations 2003–2018. ↩

18. On post-2018 economic indicators, see “Abiy Ahmed’s Hegemonistic Prosperity Party,” Borkena, June 2026; IMF and World Bank Ethiopia reports. ↩

19. Meles’s death announcement, August 2012; the cause has never been officially disclosed. ↩