The House of Federation and the Missing Constitutional Court

Thesis

The 1995 Constitution does not contain a constitutional court. It vests constitutional interpretation in the House of Federation, the upper chamber of parliament, supported by a Council of Constitutional Inquiry that screens questions before they reach the House.¹ This structural choice — to locate constitutional review in a political body rather than in a judicial one — is the single most consequential institutional decision in the 1995 settlement, and it is the structural feature that, more than any other, has determined the operational weakness of constitutional review in Ethiopia.

The article’s argument is that the House of Federation as constitutional adjudicator is a contradiction in terms. The House is a representative body whose members are selected by regional councils dominated by the ruling party; it does not have the institutional incentives to constrain the executive that the Madisonian framework requires; it does not have the institutional capacity to develop a body of constitutional jurisprudence that a judicial body would develop; and its operational record across three decades confirms what the design predicts. The House has not, in the 1995 Constitution’s history, struck down significant federal legislation as unconstitutional. It has, when called upon by the executive to issue politically consequential interpretations, issued the interpretations the executive needed. The June 2020 election-postponement interpretation is the most discussed case but is not an isolated one. The missing constitutional court is not a minor omission. It is the structural feature that ensures the constitution will remain a hollow pillar.

The text

The 1995 Constitution’s provisions on constitutional interpretation are scattered across three articles.

Article 62 lists the powers of the House of Federation. Among them, Article 62(1) gives the House the power “to interpret the Constitution.”²

Article 83 specifies the procedure: “All constitutional disputes shall be decided by the House of the Federation. The House of the Federation shall, within thirty days of receipt, decide a constitutional dispute submitted to it by the Council of Constitutional Inquiry.”³

Article 84 establishes the Council of Constitutional Inquiry as a screening body. Its eleven members include the president of the Federal Supreme Court (chair), the vice-president of the Federal Supreme Court, six legal experts appointed by the President of the Republic on recommendation of the House of Peoples’ Representatives, and three persons designated by the House of Federation from among its own members. The Council “shall have the power to investigate constitutional disputes” and “shall submit its recommendations to the House of the Federation.”⁴

The structure is significant. Constitutional review reaches the deciding body (the House of Federation) only through a screening body (the Council of Constitutional Inquiry) whose composition is significantly weighted toward executive appointment. The deciding body itself is a political chamber elected by regional councils dominated by the ruling party. There is no point in the process at which an independent judicial body has the final word.

The structural problem

Why does this matter? Constitutional review is structurally different from other adjudication in a constitutional democracy. Its purpose is to constrain the executive and legislature when they exceed constitutional bounds. The body conducting the review must therefore be institutionally insulated from the executive and legislature it is reviewing. If the reviewing body is dependent on the executive and legislature for its existence, composition, and operation, the review will not constrain them; it will rationalise them.

This is not an abstract concern. It is the lesson of every successful constitutional democracy in the post-1945 world. The U.S. Supreme Court’s strength as a constitutional reviewer rests on life tenure, financial independence, and a tradition of judicial review going back to Marbury v. Madison (1803). The German Federal Constitutional Court is selected through a procedure designed to require cross-party consensus and serves twelve-year non-renewable terms. The Indian Supreme Court has security of tenure and a tradition of judicial review developed through the Kesavananda Bharati case (1973) and after. The South African Constitutional Court has security of tenure and a strong appointment process that requires consultation with multiple constitutional institutions.⁵ In each case, the institutional design provides the reviewing body with insulation from the political branches it reviews.

The 1995 Ethiopian design does the opposite. The House of Federation is constituted by regional councils that are politically aligned with the federal ruling party (almost without exception through 2026); its members serve at the pleasure of their regional councils, which can recall them; its decisions on constitutional matters are made by political vote rather than by judicial reasoning; and it has no professional staff comparable to a constitutional court’s law clerks and research staff. The Council of Constitutional Inquiry, which screens questions before they reach the House, is itself composed of judges who serve at the pleasure of the executive and political appointees with no security of tenure. The structural conditions for independent review are absent at every stage.

Why the framers made this choice

The 1995 framers were aware of the alternative design — a constitutional court — and chose the House of Federation deliberately. The reasons cited at the time, and subsequently by Constitutional Commission members in their writings, were several.

First, the framers wanted constitutional interpretation to express the sovereignty of the constituent subjects (the nations/nationalities/peoples). A constitutional court of professional judges would not, on this view, express that sovereignty as directly as a body of representatives selected by the regions where the nationalities reside. The choice was a logical consequence of Article 8 (sovereignty in the peoples) applied to the institutional design of constitutional review.⁶

Second, the framers were concerned about the institutional capacity of the Ethiopian judiciary in 1995. The judiciary had operated under the Derg’s framework; many of its members had served under that framework; the framers did not trust it as the guardian of the new constitution. The House of Federation, composed of regional representatives who were broadly supportive of the new dispensation, was understood as a more politically reliable steward.⁷

Third, the framers cited the cost and complexity of building a separate constitutional court from scratch. The Council of Constitutional Inquiry was, in this framing, a transitional arrangement that piggybacked on the existing judicial structure (Supreme Court judges serving on the Council) while preserving political control over the final outcome.⁸

Each of these reasons has some plausibility. None of them, examined three decades later, justifies the structural pathology the choice has produced.

What the structure has produced

The House of Federation’s operational record on constitutional interpretation is the empirical evidence of the structural problem.

Volume. In the constitution’s first quarter-century, the House of Federation handled relatively few constitutional cases. The Council of Constitutional Inquiry has issued recommendations on a modest number of cases per year. Most of the cases involve regional-federal jurisdictional disputes or interpretive questions about specific articles. Very few involve challenges to federal legislation, executive action, or rights violations against the federal government.⁹

Pattern. When the House has decided politically consequential cases, the decisions have aligned with executive preferences. The June 2020 election-postponement decision is the most discussed: the House decided, on a 30-day timeline that the executive had requested, to extend the terms of all federal and regional legislatures by an act of constitutional interpretation that opposition parties and many constitutional lawyers regarded as creating new constitutional language under the guise of interpretation.¹⁰ The September 2020 declaration of the Tigray regional election as unconstitutional, the various decisions about federal-regional jurisdictional disputes, and the 2026 removal of five constituencies from Tigrayan oversight — each of these has aligned with the federal executive’s position. The House has not, in any politically consequential case, ruled against the executive.

Jurisprudence. The House has not developed a body of constitutional jurisprudence comparable to what constitutional courts in other countries have developed. There is no Ethiopian equivalent of Kesavananda (basic structure doctrine, India), Bertelsmann (proportionality, Germany), Glenister (corruption investigation independence, South Africa), or Citizens United (campaign finance, U.S.). The House’s decisions are typically short, narrowly focused on the specific question, and do not generate doctrines that constrain future executive action.¹¹ The absence of constitutional jurisprudence is significant: working constitutional democracies are constrained by the body of doctrine that constitutional courts have built up over time, even when individual decisions are politically inconvenient. Ethiopia does not have that body of doctrine because the body that would build it does not exist.

Capacity. The House of Federation lacks the professional infrastructure that constitutional adjudication requires. It does not have a research staff specialised in constitutional law. Its members are not, as a rule, lawyers; they are regional representatives whose other duties absorb most of their attention. The Council of Constitutional Inquiry, which has more legal expertise, has only screening authority. The structural capacity for serious constitutional review does not exist within the institutions the constitution has constructed.

The 2020 case in detail

The June 2020 House of Federation decision deserves examination because it illustrates the structural problem in its sharpest form.

The COVID-19 pandemic produced, in March 2020, the postponement of the August 2020 federal and regional elections by the National Electoral Board of Ethiopia. The constitution’s text provided no mechanism for election postponement. The terms of all federal and regional legislatures were due to expire in October 2020; the question was what constitutional framework would apply if no new elections had been held by that point.¹²

The federal government, through the Deputy Attorney General, presented four options to the House of Peoples’ Representatives in April 2020: (a) caretaker government; (b) declaration of state of emergency; (c) constitutional amendment; (d) constitutional interpretation by the House of Federation. The four options had different constitutional weights, with constitutional amendment (option c) being the most procedurally legitimate and constitutional interpretation (option d) being the most discretionary.¹³

The federal government recommended option (d). The House of Peoples’ Representatives, controlled by the Prosperity Party, voted to refer the matter to the Council of Constitutional Inquiry, despite objections from 25 members (mainly from the TPLF). The Council of Constitutional Inquiry, on a timeline of approximately six weeks, recommended that the constitution should be interpreted to extend the terms of all federal and regional legislatures until elections could be held within 9 months but not later than one year after the pandemic was controlled. The House of Federation, on 11 June 2020, adopted this recommendation.¹⁴

The constitutional difficulty was that the constitution’s text contained no provision for term extension. The interpretation produced an extension where the text did not provide for one. Constitutional lawyers writing about the decision noted that the reasoning was, in the words of one critic, “fallacious, self-serving and deeply flawed,” amounting to the Council “basically saying that ‘the Constitution must be interpreted this way because I said so!’” — a rewriting of the constitution under the guise of interpretation.¹⁵

The TPLF rejected the decision as unconstitutional, held its own regional election in September 2020 in defiance of the federal postponement, and produced — three months later — the November 2020 outbreak of the Tigray war. The House of Federation declared the Tigray election unconstitutional on 5 September 2020.

The structural problem the case illustrates is not that the House made the wrong decision (although many constitutional lawyers concluded that it did). It is that the House had no institutional capacity to make a different decision than the one the executive needed. The Council that screened the question was composed of executive appointees. The House that decided the question was composed of regional representatives dependent on regional governments that were Prosperity Party-controlled. The 30-day decision timeline did not permit the kind of deliberative review a constitutional court would conduct. The institutional design produced the predictable result.

What a constitutional court would have done

It is useful to imagine what an Ethiopian constitutional court — modelled on, say, the South African Constitutional Court — would have done in the 2020 case.

Such a court would have had professional law clerks who would have drafted detailed memoranda on the constitutional question. It would have held oral hearings at which all parties — federal government, opposition parties, the TPLF, civil society organisations — could have made submissions. It would have deliberated over months rather than weeks. It would have published a written decision with reasoned arguments addressing each major submission. The decision might still have been the same: extending the legislative terms could plausibly be defended under emergency-doctrine reasoning, and a constitutional court might well have approved it. But the process would have generated a record, the reasoning would have been visible, the dissents would have been published, and the next constitutional court considering a similar question would have had the precedent to work from.

The House of Federation produced none of these features. The decision was brief, the reasoning thin, the dissents not separately published, the record largely closed to outside review. Whatever the merits of the substantive outcome, the procedural failure was complete.

This is the cost of having a constitutional review body that is not a constitutional court. Even when the body reaches the right outcome (which it sometimes does), it does not generate the institutional production — jurisprudence, transparency, precedent, doctrinal development — that makes constitutional review systematically constraining.

Steelman of the defenders

The defenders of the 1995 Constitution’s design make two arguments that deserve engagement.

The first is that constitutional adjudication by a political body — rather than by a judicial one — better reflects the democratic character of the constitution, because constitutional decisions are inherently political and should be made by political representatives rather than by judges with no democratic mandate.¹⁶ This is a version of the “counter-majoritarian difficulty” argument familiar from American constitutional theory. It has weight. The reply is that the same logic should produce stronger democratic mandates for the reviewing body than the House of Federation has. Members elected by regional councils, themselves elected by the ruling party, do not have a strong democratic mandate; they have a ruling-party mandate. The House of Federation is not a more democratic body than a constitutional court would be in a healthy democracy; it is a more politically controlled body.

The second is that the House of Federation has, in fact, produced acceptable constitutional outcomes across most of its history, and that the criticism of its 2020 decision is a single-case complaint that does not generalise. The pre-2020 record, on this view, includes acceptable handling of the Silte case (2000), the Wolayta and Hadiya nationality cases, the Sidama nationality question, various federal-regional disputes — and these were handled adequately by the House.¹⁷ The reply is that “adequately” by what standard. The cases the defenders cite are cases in which the House gave the regime what it wanted in lower-stakes situations. The 2020 case is the one in which the institutional structure was tested, and it produced what the structural critique predicts. One should evaluate the institution at its hardest moments, not its easiest ones.

Multi-causal accounting

The weakness of constitutional review in Ethiopia is not solely attributable to the House of Federation design. The weakness of the judiciary as a whole, the limited legal-aid infrastructure, the absence of strong civil society organisations capable of bringing strategic litigation, the constitutional illiteracy of the general population, and the political dependence of the legal profession on the state — all contribute. The institutional design is one factor among several. But it is the factor the framers chose, and it is the factor that has continued to produce predictable results across three decades of operation.

What the article establishes

The absence of a constitutional court is the single most consequential institutional choice in the 1995 settlement. The House of Federation cannot perform the constitutional-review function that a constitutional court would perform, because it lacks the institutional insulation, professional capacity, and incentive structure that constitutional review requires. The operational record across three decades confirms the structural prediction. The constitution’s hollowness, diagnosed in the previous article, is significantly attributable to the absence of the constitutional institution that would have addressed it.

The next article takes up the federation’s operation in practice during its first decade (1995–2005), as the new constitutional order was tested by the realities it had been designed to manage.

---

References

Footnotes

1. 1995 Constitution, Articles 62, 83, 84. ↩

2. Article 62(1) of the 1995 Constitution. ↩

3. Article 83 of the 1995 Constitution. ↩

4. Article 84 of the 1995 Constitution. ↩

5. For comparative constitutional review design, see Ran Hirschl, Towards Juristocracy: The Origins and Consequences of the New Constitutionalism (Cambridge, MA: Harvard University Press, 2004); Tom Ginsburg, Judicial Review in New Democracies (Cambridge: Cambridge University Press, 2003). ↩

6. On the framers’ rationale, see Fasil Nahum, Constitution for a Nation of Nations, ch. 6; Tsegaye Regassa, “The Making and Legitimacy of the Ethiopian Constitution,” Afrika Focus 23, no. 1 (2010). ↩

7. On the framers’ distrust of the inherited judiciary, see Andreas Eshete’s various early writings. ↩

8. On the transitional-arrangement framing, see Fasil Nahum, Constitution for a Nation of Nations. ↩

9. On the volume and pattern of House of Federation cases, see Getachew Assefa, “All About Words: Discovering the Intention of the Makers of the Ethiopian Constitution on the Scope and Meaning of Constitutional Interpretation,” Journal of Ethiopian Law 24, no. 2 (2010); Adem Kassie Abebe, various writings. ↩

10. House of Federation decision, 11 June 2020. ↩

11. On the absence of constitutional jurisprudence development, see Berihun Adugna Gebeye, Constitutionalism in Ethiopia: Theoretical and Practical Issues (Edward Elgar, 2021). ↩

12. On the 2020 postponement and constitutional question, see Marew Abebe and Mekonnen Yoseph, “The Constitutionality of Election Postponement in Ethiopia amidst COVID-19 Pandemic,” Jimma University Journal of Law (2021). ↩

13. The four options discussion is documented in Marew and Mekonnen, “Constitutionality of Election Postponement.” ↩

14. House of Federation decision, 11 June 2020; see “COVID-19 and elections in Ethiopia,” SciELO, https://scielo.org.za/scielo.php?pid=S2077-49072021000100003&script=sci_arttext. ↩

15. “Constitutional Silence on Election Postponement in Ethiopia,” LexForti, January 2021, https://lexforti.com/legal-news/election-postponement-in-ethiopia/. ↩

16. This is a version of the counter-majoritarian argument; see Andreas Eshete’s writings for an Ethiopian application. ↩

17. For the institutional defence with reference to pre-2020 cases, see Tsegaye Regassa, “Sub-National Constitutions in Ethiopia,” Mizan Law Review 3, no. 1 (2009). ↩