How Others Solved the Problem: South Africa, India, Indonesia, Nigeria

Thesis

The Ethiopian constitutional pathology this series has been examining is not unique. Other post-conflict, multi-ethnic, large-population, developing-country federations have faced structurally similar problems and have, with varying degrees of success, addressed them through constitutional design and institutional construction. Four cases are particularly instructive because they share with Ethiopia some combination of post-conflict transition, multi-ethnic composition, federal or quasi-federal structure, and developmental-state ambition: South Africa from 1996, India from 1950, Indonesia after 1998, and Nigeria’s Fourth Republic from 1999. None of these is a constitutional success in absolute terms. Each is, in significant respects, an institutional failure. But each has produced constitutional institutions that the Ethiopian framework has not produced, has constrained executives in ways the Ethiopian framework has not constrained, and has developed institutional capacity for constitutional adjudication that the Ethiopian framework has not developed.¹

The article’s argument is that the constitutional cure for Ethiopia’s pathology is not unknown. It has been implemented, at scale, in countries comparable to Ethiopia. The cure has not been implemented in Ethiopia not because the cure has not been discovered but because the political will to implement it has been absent. The comparators establish what the Madisonian-Hayekian-Acemoglu-Robinson framework can accomplish when sufficiently committed political coalitions adopt it; the Ethiopian record establishes what happens when the framework is not adopted at sufficient scale. The contrast is the closing argument of the series.

South Africa, 1996

The South African transition from apartheid to constitutional democracy between 1990 and 1996 is the most directly comparable case for Ethiopia. South Africa in 1990 was a society with deep ethnic and racial divisions, severe political polarisation, a recent history of state-sponsored violence, an authoritarian government negotiating with armed opposition movements, and an international context that limited but did not predetermine the transition outcomes.² The constitutional transition that emerged from the 1990–1996 process produced:

The 1993 Interim Constitution and the Constitutional Court. The interim constitution that governed South Africa from April 1994 to February 1997 was the product of multi-party negotiations at the Convention for a Democratic South Africa (CODESA) and the Multi-Party Negotiating Process. It included 34 “Constitutional Principles” that the eventual final constitution had to satisfy, and it established the Constitutional Court as the apex constitutional review body.³ The Constitutional Court’s first major task was to certify whether the 1996 draft final constitution complied with the constitutional principles — which it did in two stages, requiring amendments before final certification.⁴

The 1996 Final Constitution. Adopted by the Constitutional Assembly (the elected parliament sitting in constituent capacity) on 8 May 1996, certified by the Constitutional Court on 4 December 1996, and entered into force on 4 February 1997. The text contains 14 chapters and 243 sections. It is, by international comparison, one of the most carefully drafted modern constitutions.⁵

The Bill of Rights (Chapter 2). A comprehensive enumeration of fundamental rights, including socio-economic rights (housing, health care, food, water, social security, education) that are justiciable. The Constitutional Court has, since 1996, developed an extensive jurisprudence on rights enforcement.⁶

The federal-provincial structure. Nine provinces with substantial autonomy, but operating within a framework that subordinates provincial powers to constitutional supremacy and federal coordination through the National Council of Provinces (the second chamber of parliament).

The Constitutional Court. Eleven justices appointed by the President from a list provided by the Judicial Service Commission, serving twelve-year non-renewable terms or to age 70.⁷ The Court’s institutional design provides genuine insulation from executive control: the appointment process requires consultation with the Chief Justice, the leaders of opposition parties, and the Judicial Service Commission. The Court has, since 1996, struck down significant legislation, including the death penalty (Makwanyane, 1995), the criminal-procedure provisions of the Riotous Assemblies Act, the President’s executive authority in various contexts, and the corruption-protection provisions involving the Hawks (Glenister, 2011).⁸

The Truth and Reconciliation Commission. Established by the Promotion of National Unity and Reconciliation Act, 1995. The TRC conducted hearings from 1996 to 1998 and produced its final report in 1998 (with the conclusion of amnesty proceedings in 2003). The TRC was not a perfect institution and its outcomes are contested, but it represented a serious institutional attempt at transitional justice that no Ethiopian process has matched.⁹

The South African outcome since 1996 has been mixed. The Constitutional Court has retained its institutional standing and continues to constrain executive action. The bill of rights has produced substantive case law on housing, health, and other socio-economic questions (TAC v. Minister of Health, Grootboom, Soobramoney). The federal-provincial structure has operated. Elections have produced peaceful transfers of power within the African National Congress’s continuing dominance (Mbeki to Motlanthe to Zuma to Ramaphosa, all through party processes; the 2024 GNU producing the ANC’s first loss of outright majority). State capture under Zuma (2009–2018) produced significant institutional damage that the post-2018 administration has been working to address.¹⁰ South Africa is not a constitutional success story without qualifications; it is a constitutional functioning system that has experienced periods of severe institutional stress.

The Ethiopian comparison is direct. South Africa in 1990 had problems Ethiopia in 1991 did not have (apartheid’s racial hierarchy, the scale of armed opposition’s military capacity, the international sanctions framework). Ethiopia in 1991 had problems South Africa in 1990 did not have (the immediately following Eritrean independence, the nationality politics of multiple regions, the lack of any established business class capable of negotiating). But the broad parameters — post-conflict transition, multi-ethnic society, federal structure, developmental-state ambition — were sufficiently similar to make the comparison meaningful.

South Africa built a constitutional court. Ethiopia did not. South Africa built a justiciable bill of rights. Ethiopia put one in its constitutional text but did not build the institutional infrastructure to enforce it. South Africa established a truth commission. Ethiopia did not (the Special Prosecutor’s Office for Derg-era crimes produced trials but not a comparable national process). The institutional choices South Africa made are not unknown to Ethiopian constitutional drafters; they were available. They were not made.

India, 1950

The Indian Constitution that entered into force on 26 January 1950 is the longest-functioning constitution of any large developing country.¹¹ India in 1950 was a society of comparable demographic complexity to Ethiopia (a population then of approximately 360 million, speaking dozens of languages, organised across multiple religious and caste lines, emerging from partition violence that had killed perhaps a million people and displaced fifteen million more). The constitutional construction that managed this complexity has, despite the 1975–1977 Emergency under Indira Gandhi and the periodic strains of communal violence, secessionist movements, and Maoist insurgencies, retained its core institutional character for seventy-five years.

The institutional features of the Indian constitutional order most directly relevant to the Ethiopian comparison:

The Supreme Court. A court of constitutional review with security of tenure for justices (appointed by the President on advice of the Chief Justice, serving to age 65). The Court has developed, since the 1950s, an extensive constitutional jurisprudence including the “basic structure” doctrine established in Kesavananda Bharati (1973), which holds that parliament cannot, even by constitutional amendment, alter the basic structure of the constitution.¹² The basic structure doctrine has been used to strike down constitutional amendments (most famously in 1980, against the Emergency-era amendments).

Federalism without ethnic territorialism. India is organised into 28 states and eight union territories, structured primarily by linguistic and historical considerations rather than by ethnic identity. The States Reorganisation Act of 1956 produced the major linguistic reorganisation, and subsequent additions (Telangana from Andhra Pradesh, the northeastern states) have followed similar processes. The Indian states do not have constitutional right of secession; the constitution explicitly forbids secession (though India has, since independence, experienced secessionist insurgencies in Kashmir, the Northeast, and Punjab that it has managed through a combination of military, political, and constitutional means).¹³

A directly elected lower house (Lok Sabha) and indirectly elected upper house (Rajya Sabha). The institutional structure provides for federalism without the ethnic-territorial subordination that the Ethiopian constitution’s Article 8 and Article 39 produce.

The Election Commission of India. A constitutional body responsible for conducting national and state elections, headed by a Chief Election Commissioner appointed for fixed terms with security of tenure. The Election Commission has, despite periodic controversies, produced electoral processes that have included multiple changes of government at both national and state levels.¹⁴

The Comptroller and Auditor General. A constitutional body that audits all government accounts and reports to parliament. The CAG has been institutionally significant in producing accountability for corruption and mismanagement.

The Indian record is not unblemished. The 1975–1977 Emergency was a major constitutional rupture, during which fundamental rights were suspended, opposition leaders were imprisoned, the press was censored, and a forced sterilisation programme produced widespread documented abuses. The post-Emergency political response — Indira Gandhi’s electoral defeat in 1977, the establishment of the Janata Party government, the subsequent restoration of constitutional ordinary operation — demonstrated that the institutional framework could survive even substantial executive overreach.¹⁵ India has continuing problems: communal violence, the constitutional status of Kashmir after the August 2019 abrogation of Article 370, the institutional pressures on judicial independence and press freedom under the post-2014 Modi government. But the constitutional architecture has continued to function and to constrain in important domains.

The Ethiopian comparison establishes that a multi-ethnic, multi-religious, post-colonial federation can produce sustained constitutional functioning over decades with institutional features (constitutional court, election commission, justiciable rights chapter) that Ethiopia did not build. The seventy-five-year demonstration is not theoretically available; it is empirically established.

Indonesia after 1998

The Indonesian transition from the Suharto authoritarianism (1966–1998) to constitutional democracy is the most recent of the comparators and the one most directly comparable to Ethiopia in timing. Indonesia in 1998 was a multi-ethnic, multi-religious, archipelagic country with a population of approximately 210 million, a developmental-state legacy from Suharto’s New Order, a recent history of armed conflict in East Timor and other regions, and the financial crisis of 1997–98 that had destroyed the legitimacy of the existing regime.¹⁶

The Indonesian constitutional transition produced:

Four constitutional amendments, 1999–2002. The original 1945 Constitution was amended in four rounds to introduce direct presidential elections, term limits for the presidency (two five-year terms), an independent constitutional court, a strengthened legislature, and an expanded bill of rights.¹⁷

The Constitutional Court (Mahkamah Konstitusi), 2003. Nine justices, appointed three each by the President, the parliament (DPR), and the Supreme Court, serving five-year terms renewable once. The Court has jurisdiction over constitutional review of legislation, dispute resolution between state institutions, dissolution of political parties, and dispute resolution over election results. The Court has, since its establishment, struck down significant legislation and adjudicated major political disputes.¹⁸

Direct presidential elections. First held in 2004, producing a peaceful transfer of power from Megawati Sukarnoputri to Susilo Bambang Yudhoyono. The 2014 election produced a transfer to Joko Widodo from outside the established political elite. Subsequent elections in 2019 and 2024 have continued the pattern of competitive elections producing alternation through electoral process.¹⁹

Aceh special autonomy and the 2005 peace agreement. The Free Aceh Movement (GAM) insurgency that had been active since 1976 was resolved through the August 2005 Memorandum of Understanding between the Indonesian government and GAM, mediated by former Finnish president Martti Ahtisaari. The MoU produced a comprehensive autonomy framework for Aceh under the Indonesian constitutional structure, including a regional government, local political parties, sharia law for Muslims, and revenue-sharing arrangements.²⁰ Aceh has since 2005 been integrated into the Indonesian federal structure without producing further armed conflict.

Decentralisation. Substantial administrative and fiscal decentralisation to provinces, regencies, and municipalities since 1999. The decentralisation has not been organised primarily around ethnic-territorial principles; it has produced administrative units with diverse populations governed through local elections.

The Indonesian record is mixed. The post-2014 trajectory under Joko Widodo and his successors has produced some institutional erosion: the 2023 Constitutional Court ruling that effectively created a path for Widodo’s son to qualify for the 2024 vice-presidency, the various rule-of-law concerns under the Prabowo Subianto government elected in 2024.²¹ But the basic constitutional functioning has been maintained: elections occur on schedule, the Constitutional Court continues to operate, the federal structure functions, the Aceh integration has held.

The Ethiopian comparison is particularly direct. Indonesia in 1998 and Ethiopia in 1991 were comparable in many ways: large population, multi-ethnic, developmental-state legacy, post-conflict, archipelagic/regional complexity. Indonesia produced sustained constitutional functioning that Ethiopia did not. The constitutional court that Indonesia built in 2003 is, structurally, the kind of institution that the 1995 Ethiopian framers chose not to build.

Nigeria’s Fourth Republic

Nigeria’s Fourth Republic, since 1999, is the most institutionally troubled of the four comparators but is also the most directly comparable to Ethiopia in respect of ethnic complexity, federal structure, and developmental challenges. Nigeria has 36 states and a Federal Capital Territory, three major ethnic groups (Hausa-Fulani, Yoruba, Igbo) and over 250 smaller ethnic groups, a population now exceeding 230 million, and a recent history including the 1967–1970 Biafran civil war (one to three million dead) and multiple military coups.²²

The Nigerian constitutional order that has operated since 1999 includes:

The 1999 Constitution. A presidential federal constitution with a directly elected president, a bicameral National Assembly, and 36 state governments with directly elected governors. The constitution was drafted under military supervision and has been the subject of continuing reform discussions, but it has provided the framework for Nigerian governance for over 25 years.²³

Supreme Court of Nigeria. A court of constitutional review that has, in significant cases, adjudicated major political disputes (the 2003 and 2007 election disputes, gubernatorial removals, the constitutional status of various legislative actions). The Supreme Court’s independence has been periodically contested but has been maintained in important respects.²⁴

Elections producing alternation. The 2015 election produced the first electoral transfer of power from an incumbent president to an opposition candidate in Nigerian history (Muhammadu Buhari defeating Goodluck Jonathan). The 2023 election produced a continuing competitive process (Bola Tinubu winning a three-way contested election). The constitutional electoral mechanism has produced power transfer through electoral process, although with significant procedural and substantive problems.

Federal-state revenue allocation. A constitutional mechanism for sharing oil revenues among federal, state, and local governments that has, despite massive corruption, provided the framework for state-level governance.

Sharia law in northern states. Twelve northern states have implemented sharia legal systems since 1999, operating in parallel with the federal constitutional framework. The arrangement has produced significant practical problems but has, in operational terms, accommodated religious-legal pluralism within the federal structure.²⁵

The Nigerian record includes substantial problems: the Boko Haram insurgency (2009–present), the secessionist tensions in the southeast (the Indigenous People of Biafra movement), the banditry crisis in northern states, ethnic-religious violence in the Middle Belt, and pervasive corruption. The constitutional architecture has constrained executive power imperfectly: the Buhari and Tinubu governments have engaged in various constitutionally questionable practices. But the basic architecture has functioned in ways the Ethiopian architecture has not. Elections occur. Power transfers occur. The Supreme Court adjudicates. The federation persists despite severe stress.

The Ethiopian comparison is, in some ways, the most useful of the four because Nigeria’s institutional functioning is the most imperfect. If Nigeria’s Fourth Republic, with all its problems, has produced constitutional functioning that Ethiopia has not, the implication is that the threshold for constitutional functioning is achievable for Ethiopia. The bar is not set at South African or Indian institutional sophistication; it is set at Nigerian functioning, and Nigerian functioning is achievable.

What the comparators establish

The four cases establish, collectively, four propositions.

First, constitutional functioning is possible in conditions comparable to Ethiopia’s. None of the comparators had ideal conditions for constitutional construction. Each faced severe ethnic, religious, regional, or post-conflict challenges. Each built institutional infrastructure that produced sustained constitutional operation.

Second, the specific institutional features Ethiopia lacks are achievable. Constitutional courts, independent electoral commissions, justiciable rights chapters, judicial security of tenure — these are not exotic institutions confined to Western democracies. They have been built in the comparators and have functioned, with imperfections, for decades.

Third, the specific design choices Ethiopia made are not necessary consequences of its conditions. Ethnic-territorial federalism with constitutional secession provisions, constitutional review by a political body, state land ownership — these are choices, not inevitabilities. The comparators chose differently and produced different outcomes.

Fourth, the political-economy conditions for institutional construction are constructible. Each comparator’s institutional construction was a political achievement, not an automatic consequence of social conditions. The political coalitions that demanded the institutional features assembled themselves through processes that are, in significant ways, replicable.

The cure for the Ethiopian constitutional pathology is known. It has been implemented elsewhere. The obstacle has not been knowledge but political will.

Steelman of opposing readings

The objection that this article must engage is that the comparators are not actually comparable, that each operated in conditions sufficiently distinct from Ethiopia’s that the lessons do not transfer.

The specific objections:

(a) South Africa had the apartheid legacy that produced a unique negotiated transition with substantial international support.

(b) India had the British colonial inheritance of common-law institutions and a Western-trained constitutional drafting team.

(c) Indonesia had the New Order’s developmental-state infrastructure and the 1997 financial crisis as a circumstantial accelerator.

(d) Nigeria had oil revenues that financed institutional construction.

Each objection has weight. The reply is that each comparator also lacked things Ethiopia had (Ethiopia had a more compact territory than India or Nigeria; Ethiopia had its independence preserved through the colonial period and so retained sovereign continuity South Africa did not; Ethiopia had a more religiously unified population than Indonesia or Nigeria). The differences cut in multiple directions. The comparisons are not perfect, but they are not therefore not comparisons. The institutional choices Ethiopia made were choices made among options that the comparators show were available.

Multi-causal accounting

Each comparator’s outcomes reflect multiple causes; none of the cases is a pure constitutional-design success or failure. South Africa’s outcomes reflect the ANC’s organisational character, the timing of the post-Cold War international moment, the South African business sector’s accommodation, and many other factors beyond constitutional design. India’s outcomes reflect the Congress Party tradition, the inheritance of British institutions, the post-1947 political settlement, and much else. Indonesia’s outcomes reflect the New Order’s institutional residue, the post-1998 reform coalition, the specific accommodation with the military, and other factors. Nigeria’s outcomes reflect the post-Biafra political settlement, the oil-rentier political economy, and the religious-regional balance. The constitutional design is one factor among several in each case, but it is a factor that, in each case, was deliberately constructed in ways the Ethiopian case did not match.

What the article establishes

The cure for the Ethiopian constitutional pathology is known. It has been implemented at varying scales of success in comparable countries. The institutional features Ethiopia lacks — constitutional court, independent electoral commission, justiciable rights chapter with effective enforcement, judicial security of tenure, federal structure that does not depend on ethnic-territorial subordination, electoral mechanism that produces alternation — are achievable. The Ethiopian constitutional pathology is not the result of unique conditions that prevent institutional construction; it is the result of choices that the comparators show could have been made differently.

The next article takes up the political-economy explanation for why each Ethiopian regime has, despite the comparator evidence, repeated the pattern.

---

References

Footnotes

1. For comparative constitutional design, see Tom Ginsburg and Aziz Huq, eds., Assessing Constitutional Performance (Cambridge: Cambridge University Press, 2016); Mark Tushnet, The New Constitutionalism (forthcoming). ↩

2. On the South African transition, see Heinz Klug, Constituting Democracy: Law, Globalism, and South Africa’s Political Reconstruction (Cambridge: Cambridge University Press, 2000); Patti Waldmeir, Anatomy of a Miracle: The End of Apartheid and the Birth of the New South Africa (W.W. Norton, 1997). ↩

3. Interim Constitution of the Republic of South Africa, Act 200 of 1993, including Schedule 4 (Constitutional Principles). ↩

4. In re: Certification of the Constitution of the Republic of South Africa, 1996 (CCT 23/96); In re: Certification of the Amended Text (CCT 37/96). ↩

5. Constitution of the Republic of South Africa, 1996. ↩

6. On socio-economic rights jurisprudence, see Sandra Liebenberg, Socio-Economic Rights: Adjudication under a Transformative Constitution (Juta, 2010). ↩

7. Constitution of South Africa, sections 174–178. ↩

8. S v Makwanyane 1995 (3) SA 391 (CC); Glenister v President of the Republic of South Africa 2011 (3) SA 347 (CC). ↩

9. On the TRC, see Antjie Krog, Country of My Skull (Random House, 1998); the TRC Final Report, 1998 (extended in 2003). ↩

10. On state capture, see Judicial Commission of Inquiry into Allegations of State Capture, Final Report, 2022 (the Zondo Commission Report). ↩

11. Constitution of India, entered into force 26 January 1950. ↩

12. Kesavananda Bharati v. State of Kerala AIR 1973 SC 1461. ↩

13. On Indian federalism, see Granville Austin, The Indian Constitution: Cornerstone of a Nation (Oxford: Clarendon Press, 1966); Working a Democratic Constitution: The Indian Experience (Oxford: OUP, 1999). ↩

14. On the Election Commission of India, see Madhav Khosla, India’s Founding Moment (Harvard, 2020). ↩

15. On the Emergency, see Christophe Jaffrelot and Pratinav Anil, India’s First Dictatorship: The Emergency, 1975–1977 (Hurst, 2020). ↩

16. On the Indonesian transition, see Edward Aspinall, Opposing Suharto: Compromise, Resistance, and Regime Change in Indonesia (Stanford, 2005); Marcus Mietzner, Money, Power, and Ideology (Singapore: NUS Press, 2013). ↩

17. 1945 Constitution of the Republic of Indonesia, as amended 1999, 2000, 2001, 2002. ↩

18. On the Constitutional Court, see Stefanus Hendrianto, Law and Politics of Constitutional Courts: Indonesia and the Search for Judicial Heroes (Routledge, 2018). ↩

19. On Indonesian elections, see Edward Aspinall and Marcus Mietzner, eds., Problems of Democratisation in Indonesia (ISEAS, 2010). ↩

20. On the Aceh agreement, see Edward Aspinall, Islam and Nation: Separatist Rebellion in Aceh, Indonesia (Stanford, 2009). ↩

21. On post-2014 institutional erosion, see Marcus Mietzner, “Sources of Resistance to Democratic Decline,” Asian Studies Review 44, no. 4 (2020). ↩

22. On Nigeria’s Fourth Republic, see Eghosa E. Osaghae, Crippled Giant: Nigeria Since Independence (Indiana, 1998); Wale Adebanwi, Nation as Grand Narrative: The Nigerian Press and the Politics of Meaning (Rochester, 2016). ↩

23. 1999 Constitution of the Federal Republic of Nigeria. ↩

24. On Supreme Court of Nigeria, see Hakeem Yusuf, Transitional Justice, Judicial Accountability and the Rule of Law (Routledge, 2010). ↩

25. On sharia in northern Nigeria, see Brandon Kendhammer, Muslims Talking Politics (University of Chicago, 2016). ↩