The Federation in Practice: The First Decade, 1995–2005

Thesis

The first decade of the 1995 Constitution’s operation, from August 1995 to May 2005, is the period in which the constitutional order’s basic operating pattern was established. It is the period of TPLF/EPRDF consolidation under Meles Zenawi as Prime Minister, the war with Eritrea (1998–2000), the post-2001 internal TPLF crisis, and the run-up to the 2005 election that would be the constitution’s first severe test. The decade is often treated, by sympathetic observers, as a period of relatively functional ethnic federalism — the constitution settling into operation, regional governments forming and operating, language rights being implemented, electoral processes running on their nominal cycles.¹ By less sympathetic observers it is treated as the period of consolidation of single-party authoritarianism under constitutional cover.²

The article’s argument is that both readings are partly right. The first decade did produce real institutional output that the Derg years had not produced, and that comparison with what came after 2005 makes look retrospectively peaceful. The first decade also operated within a single-party framework that prevented the constitutional architecture from being tested at its load-bearing points. The federation functioned in the conditions for which it was designed — EPRDF dominance, peripheral opposition, ethnically organised regional politics under coordinated central control. It was not yet asked to function in the conditions that would test it: serious electoral competition, intra-coalition rupture, federal-regional conflict on existential questions. Those tests would come later. The pre-2005 record is, in this sense, an evaluation of the constitutional order under favourable conditions.

The political shape of the decade

The decade was politically structured by EPRDF dominance. The 1995 general election, the first under the new constitution, returned an EPRDF supermajority in the House of Peoples’ Representatives. The OLF, AAPO, and several smaller opposition parties had boycotted; the field was effectively non-competitive. The 2000 general election produced a similar outcome under similar conditions, although with some opposition participation.³

Meles Zenawi was Prime Minister throughout the decade (1995–2012). He had been chairman of the EPRDF and effective leader of the transition since 1991. The 1995 Constitution made the Prime Minister the head of government in a parliamentary system; the EPRDF’s control of parliament made Meles’s control of the executive secure for the duration. The President of the Federal Democratic Republic was, by constitutional design, a largely ceremonial figure — Negaso Gidada (1995–2001) and then Girma Wolde-Giorgis (2001–2013). Effective political power was concentrated in the Prime Minister’s office and the EPRDF Executive Committee.

The EPRDF was itself a coalition of four constituent parties: the Tigray People’s Liberation Front, the Amhara National Democratic Movement (renamed from EPDM), the Oromo People’s Democratic Organisation, and the Southern Ethiopian People’s Democratic Movement.⁴ Each had primary responsibility for its corresponding regional state, and each was structured around the political logic of the nationality it claimed to represent. The federation’s nominal regional autonomy operated, in practice, through the coordinated decisions of EPRDF constituent parties whose senior leadership all sat together in the EPRDF Executive Committee. Federalism in form, single-party coordination in substance.

This is the operational arrangement that mattered. The 1995 Constitution’s federal structure was a layer on top of an EPRDF political structure. The political structure determined what federalism could actually mean. Regional governments could exercise the autonomy the constitution granted them as long as the exercise was consistent with EPRDF coalition policy; when it was not consistent, the EPRDF mechanisms of coordination produced compliance. The institution that the constitution had imagined — the regional government as a constitutional unit with its own democratic legitimacy and federal-relations standing — operated, in practice, as the regional affiliate of a national coalition.

What functioned

Several things did function in the first decade.

Regional administration in working languages. The constitution’s commitment to regional working languages was implemented. Tigrinya, Afaan Oromo, Somali, Afar, and various languages of the SNNPR became the working languages of their respective regional administrations. Primary education was made available in regional languages. The pre-1991 dominance of Amharic as the sole language of the modern state was diminished in practice as well as in principle.⁵

Federal-state revenue-sharing. The 1995 Constitution’s Article 96 vested taxing powers in the federal government, the states, and concurrently, with revenue-sharing arrangements specified by federal proclamation. The revenue-sharing system operated in the first decade and produced regular transfers from the federal government to the regions. The proportionality of transfers and the dependence of regional governments on federal transfers were issues raised by some scholars — Solomon Negussie’s work on fiscal federalism documented the concerns — but the basic system functioned.⁶

Regional state councils. The regional councils elected in the 1995 and 2000 elections did meet, did pass regional laws, did adopt regional budgets, and did establish regional courts. They did so as EPRDF-affiliated bodies, but they did so. The constitutional structure of regional government was institutionalised in a way that the Derg’s autonomous regions had not been.

Electoral cycles. Elections were held on the constitutional schedule: 1995, 2000. The conduct was problematic but the cycles were maintained. The 1992 elections under the transitional government had been irregularly timed and problematic in conduct; the post-1995 elections were more regular.

Eritrean federation dissolution managed peacefully. The 1998–2000 war was a major rupture, but the political infrastructure of the post-1991 settlement — the EPRDF, the constitutional framework, the regional state structures — operated through the war and survived it without constitutional crisis. This is more than nothing.

These functioning elements are not contemptible. The decade is not a record of constitutional failure across the board. The argument of this article is that the functioning elements operated within a framework that, for the duration of the decade, was not asked to do the hard constitutional work that comes when politics becomes genuinely competitive.

What did not function

Several things did not function in the first decade, in ways that revealed the structural pathologies that subsequent decades would exhibit more dramatically.

Constitutional review. As discussed in the previous article, the House of Federation did not develop a serious constitutional jurisprudence in the first decade. The cases it handled were primarily federal-regional jurisdictional questions and nationality identity questions; it did not produce the kind of constitutional rights jurisprudence that would have constrained executive action. The institutional weakness was visible from the beginning.

Press freedom. The Press Proclamation of 1992 (Proclamation No. 34/1992) had established a relatively permissive framework. In operation, the framework was eroded through prosecutions, licensing restrictions, and de facto pressures throughout the decade. Independent newspapers — Addis Tribune, Ethiop, Reporter, and others — operated under continuing pressure. Journalists were detained periodically. The post-1995 constitutional commitment to free expression (Article 29) was not effectively defended by any constitutional institution.⁷

Opposition political space. The OLF, AAPO, and other parties operated under conditions that ranged from severely restricted to effectively suppressed. The OLF’s return to opposition after withdrawing from the transitional government in 1992 was treated as an insurgency rather than a political party question. The AAPO was harassed continuously. The legitimate space for organised opposition was much narrower than the constitutional text suggested.⁸

Civil society. The constitutional commitment to freedom of association (Article 31) coexisted with a Charities and Societies Proclamation (Proclamation No. 621/2009, later) and earlier administrative practices that imposed significant restrictions on civil society organisations. The decade’s civil society sector was thin, partly because of these restrictions.⁹

Judicial independence. The federal judiciary’s relationship to the executive was structurally weak. Federal Supreme Court justices were appointed by the President on the recommendation of the Prime Minister, with House of Peoples’ Representatives confirmation that was, given EPRDF dominance, a formality. The judiciary in the first decade did not develop a tradition of independent decision in politically consequential cases.¹⁰

The Eritrean war and after

The 1998–2000 war with Eritrea is the major external event of the decade and had consequences for the constitutional order’s operation that should be noted.

The war was prosecuted by the federal executive with minimal parliamentary or judicial constraint. The decision to escalate from the May 1998 Badme incident to full conventional war, the decision to mobilise hundreds of thousands of troops, the decisions about war aims and termination — these were made by the Prime Minister’s office and the EPRDF Executive Committee. The constitutional structures that, in a constitutional democracy, would have at least nominally been involved (parliamentary war powers, judicial scrutiny of emergency measures) played no operational role.

After the war, the TPLF underwent a major internal crisis in 2001. A faction of the TPLF Central Committee, led by Seyoum Mesfin, Gebru Asrat, and others, broke with Meles over war conduct and post-war strategy. The faction was expelled from the TPLF in March 2001 in what became known as the “TPLF split.”¹¹ The constitutional significance of this is twofold. First, it confirmed that ultimate political authority in Ethiopia resided in the TPLF Executive Committee, which could expel dissenting members and produce immediate consequences for them; constitutional positions could be reshuffled accordingly. Second, it removed from the TPLF leadership figures who had been advocating more cautious approaches to several questions (post-war reconstruction, internal governance, succession) and consolidated Meles’s authority within the TPLF and therefore within the EPRDF and the federal executive. The constitutional design did not regulate party politics; party politics regulated the constitutional design.

Toward 2005

By the early 2000s, several developments were producing the conditions that would test the constitutional order more seriously than the first decade had tested it.

The Ethiopian economy was growing. Post-war reconstruction, agricultural recovery, and the early stages of the export-led growth strategy that Meles would pursue more aggressively after 2005 were producing a rising urban middle class with disposable income and political consciousness.¹²

The opposition was reorganising. The Coalition for Unity and Democracy (CUD/Kinijit) was being formed from a coalition of four parties (the All Ethiopia Unity Party, the Ethiopian Democratic League, the Ethiopian Democratic Unity Party-Medhin, and the Rainbow Alliance). The United Ethiopian Democratic Forces (UEDF) was a parallel coalition forming on a more federalist platform. These coalitions were better organised, better funded, and more politically capable than the post-1995 opposition.¹³

The EPRDF was, according to several analyses, expecting an easy win in the 2005 election that would demonstrate Ethiopian democratic progress internationally. This expectation produced a relatively open campaign environment — opposition parties were granted unprecedented access to state media, allowed to hold rallies in major cities, and treated with relative restraint by security services in the pre-election period.¹⁴

The 2005 election would test whether the constitutional structure could absorb a serious opposition challenge. The next article takes up that test in detail.

Steelman of the opposing reading

Two opposing readings of the 1995–2005 decade deserve engagement.

The first is the EPRDF-sympathetic reading: that the decade was a period of state-building in extremely difficult conditions, that the constitutional structure functioned about as well as anyone had reason to expect, and that the criticism of it from later perspectives ignores the achievements (regional autonomy, language rights, the survival of the federation through the Eritrean war).¹⁵ This reading is partly right. The first decade did establish institutional structures that survived and that produced real benefits for some Ethiopians. The reply is that the structures operated under conditions that did not test them seriously, and that the operational record under serious test (2005 and after) shows the structural weakness more clearly than the pre-2005 record does.

The second is the comprehensive-critique reading: that the first decade was a period of single-party authoritarian consolidation under constitutional cover, that the apparent functioning of the federation was a veneer over EPRDF coordination, and that the constitutional structure was always intended as a legitimating device rather than as a constraint.¹⁶ This is partly right. The constitutional structure did operate within an EPRDF framework that was politically dominant. The reply is that “intended” overstates the framers’ intentions; some at least, including Andreas Eshete in his more idealistic moments, did believe the constitutional framework would produce constitutional constraint over time. The framers were wrong about what would follow, but they were not necessarily cynical about it.

Multi-causal accounting

The 1995–2005 decade’s mixed record reflects (a) the EPRDF’s effective political consolidation, which both enabled the federation’s functioning and prevented its testing; (b) the absence of organised opposition with the capacity to challenge the EPRDF until late in the decade; (c) the structural weaknesses of the constitutional institutions (House of Federation, judiciary) that meant the institutional capacity for constitutional constraint had not been built; (d) the external constraints of the Eritrean war and post-war reconstruction; and (e) the broader international context that rewarded Ethiopia’s developmental performance more than its constitutional development.

What the article establishes

The first decade of the 1995 Constitution’s operation produced real institutional output but did not test the constitutional structure at its load-bearing points. The federation functioned in the conditions for which it was designed — EPRDF dominance, peripheral opposition, ethnically organised regional politics under coordinated central control. It was not yet tested by serious electoral competition or intra-coalition rupture. The 2005 election was the first such test. The result — examined in the next article — would reveal what the first decade had concealed.

---

References

Footnotes

1. For sympathetic accounts of the period, see Yonatan Fessha, Ethnic Diversity and Federalism: Constitution Making in South Africa and Ethiopia (Aldershot: Ashgate, 2010); Solomon Negussie, Fiscal Federalism in the Ethiopian Ethnic-based Federal System (Nijmegen: Wolf Legal Publishers, 2006). ↩

2. For more critical accounts, see Marina Ottaway, “The Ethiopian Transition: Democratization or New Authoritarianism?,” Northeast African Studies 2, no. 3 (1995); Kjetil Tronvoll, Ethiopia: A New Start? (London: Minority Rights Group, 2000); Lovise Aalen, “Ethnic Federalism in a Dominant Party State: The Ethiopian Experience 1991–2000,” CMI Report 2002. ↩

3. On the 1995 and 2000 elections, see Aalen, “Ethnic Federalism in a Dominant Party State”; Tronvoll, Ethiopia: A New Start? ↩

4. On the EPRDF structure, see Sarah Vaughan, Ethnicity and Power in Ethiopia, PhD thesis, University of Edinburgh, 2003. ↩

5. On language-rights implementation, see Yonatan Fessha, Ethnic Diversity and Federalism, ch. 5. ↩

6. Solomon Negussie, Fiscal Federalism in the Ethiopian Ethnic-based Federal System. ↩

7. Press Proclamation No. 34/1992; on press freedom outcomes in the decade, see Committee to Protect Journalists Ethiopia reports 1995–2005. ↩

8. On opposition political space, see Aalen, “Ethnic Federalism in a Dominant Party State”; Lovise Aalen and Kjetil Tronvoll, “The 2008 Ethiopian Local Elections: The Return of Electoral Authoritarianism,” African Affairs 108 (2009). ↩

9. Charities and Societies Proclamation No. 621/2009. On pre-2009 civil society conditions, see Hailegabriel G. Feyissa, “The Politicization and Persistence of Repression Against Ethiopian Civil Society,” in Civil Society in Africa, ed. E. Obadare (Springer, 2014). ↩

10. On judicial independence in the period, see Getachew Assefa, “All About Words”; Adem Kassie Abebe, various writings. ↩

11. On the 2001 TPLF split, see Paulos Milkias, Haile Selassie, Western Education, and Political Revolution in Ethiopia (Youngstown, NY: Cambria Press, 2006); Medhane Tadesse and John Young, “TPLF: Reform or Decline?,” Review of African Political Economy 30, no. 97 (2003). ↩

12. On post-2000 economic growth and the rising middle class, see Christopher Adam and Stefan Dercon, “The Political Economy of Development: An Assessment,” Oxford Review of Economic Policy 25, no. 2 (2009). ↩

13. On the 2004 formation of CUD/Kinijit, see “Examining the 2005 Ethiopian Parliamentary Election Results,” John Ishiyama, https://scholarworks.wmich.edu/cgi/viewcontent.cgi?article=1117&context=africancenter_icad_archive. ↩

14. On EPRDF expectations and pre-election opening, see Medhane Tadesse and John Young analyses; “For Ethiopia’s political tribes, it’s 2005 redux,” Ethiopia Insight, May 2020, https://www.ethiopia-insight.com/2020/05/22/for-ethiopias-political-tribes-its-2005-redux/. ↩

15. This is broadly the position of EPRDF-aligned scholars and of some Ethiopian constitutional scholars who emphasise institutional continuity. ↩

16. This is the position of the comprehensive critique tradition, including Aalen, Tronvoll, and the broader human-rights-organisations literature. ↩