Ethiopia’s regions are de jure autonomous but de facto power remains centralised
Summarising his recent paper in Regional and Federal Studies, Nigussie Daba Heyi examines the real extent of regional policy autonomy within Ethiopia’s federal system, revealing a significant gap between constitutional design and political practice. Drawing on the Regional Authority Index and a mixed-methods assessment of self-rule and interactive-rule authority, he compares Ethiopia’s subnational units with other federations - and with one another. The analysis shows that while Ethiopia’s regions enjoy relatively strong de jure autonomy, their de facto influence over policymaking remains limited, constrained by fiscal dependence, centralised decision-making, and weak intergovernmental coordination. This gap between formal authority and practical power highlights the fragility of Ethiopia’s federal promise and raises important questions about the country’s evolving governance architecture.
This examination of the Ethiopian federation examines the autonomy of the country’s regional states using the Regional Authority Index (RAI), measuring both de jure and de facto autonomy and comparing Ethiopia with other federations. It also compares three sub-national units within Ethiopia. The study extends the framework and applies the RAI to assess self-rule and interactive-rule dimensions of regional policy autonomy. The quantitative analysis draws on survey scores provided by senior experts and officials between December 2021 and May 2022. Levels of autonomy were assessed separately for self-rule and interactive-rule indices, and variations across regions were examined using non-parametric Kruskal–Wallis tests.
The self-rule and interactive-rule dimensions are systematically disaggregated into sub-areas—institutional depth, policy scope, fiscal autonomy, borrowing autonomy, and representation for self-rule, and lawmaking, executive control, fiscal control, borrowing control, and constitutional reform for interactive-rule. Relevant post-constitution policy documents, laws and official reports were reviewed to code the degree of regional autonomy. Quantitative data were gathered first, followed by qualitative data, and then merged for comparative analysis. The qualitative component included around 20 interviews and a survey of 215 respondents from 17 purposively selected sectors; qualitative data were thematically coded and used to reinforce the quantitative findings (see the paper for full details of the methods used).
Oromia and the Southern Nations, Nationalities and Peoples (SNNPR) -both constituent units of the federation - and Addis Ababa, the federal capital (often treated as a regional state despite its distinct constitutional status), were selected. Oromia is a relatively central region, with a large population (38.8 per cent of the country’s total), a large land area, a largely homogeneous territorial group (the Oromo), and a recent history of extensive protests (2014–2018) demanding greater self-rule and political representation - demands that helped trigger leadership change within the EPRDF in 2018. The SNNPR, by contrast, is peripheral, medium-sized (21.1 per cent of the population) and ethnically heterogeneous, with longstanding inter-ethnic struggles for autonomy that culminated in its recent fragmentation. Addis Ababa is also significant due to its political, administrative and fiscal importance, and because it hosts numerous national and international institutions. Although the FDRE Constitution and Proclamation No. 361/2003 provide the legal basis for regional and city policy-making, these three cases illustrate how a common constitutional framework operates across very different regional contexts.
The findings show that de jure self-rule autonomy in Ethiopia is lower than in Switzerland and Belgium but higher than in India. Switzerland remains the federation with the highest level of self-rule autonomy. Within Ethiopia, de jure autonomy in institutional depth, representation and policy scope is comparatively higher than in its fiscal dimensions. However, de facto self-rule and interactive-rule autonomy are lower than their de jure counterparts. Regions often experience financial and resource constraints that limit their ability to carry out their responsibilities, and they do not co-determine federal policymaking.
Overall, Ethiopia’s de jure self-rule autonomy compares favourably with some mature federations. Yet de facto autonomy remains weaker than de jure autonomy. Ethiopia’s de jure self-rule capacity is also stronger than its de jure interactive-rule capacity, as regions do not formally participate in federal policymaking and the second chamber (the HoF) plays a limited role. Unlike the Swiss cantons or Belgian communities, Ethiopian regions lack institutionalised mechanisms for influencing federal policies and legislation. Their weak influence, combined with centralised policy practice, risks producing greater policy convergence over time.
The de facto interactive/shared-rule autonomy is also lower than its de jure equivalent. The Kruskal–Wallis test indicates a statistically significant difference among the three units in exercising self-rule autonomy: Oromia scores slightly higher, at a moderate level, while the SNNPR and Addis Ababa score low. However, there is no statistically significant difference among the units in exercising interactive/shared-rule autonomy. This suggests that all three units lack meaningful influence over federal policymaking—whether through intergovernmental relations or through the HoF, whose policy role remains weak. Intergovernmental consultations are ad hoc rather than routinised, and they do not produce binding decisions or meaningful negotiation on federal policies affecting the regions. This reflects both temporal variation (between de jure and de facto autonomy) and territorial variation within the federation.
Ultimately, regional governments in Ethiopa do not fully exercise the constitutional autonomy granted to them.
About the author: Nigussie Daba Heyi is Assistant Professor at Ethiopian Civil Service University.
This blog summarises this article from the Regional and Federal Studies journal. It represents the views of the author, and not those of Regional and Federal Studies or the University of Edinburgh.