Implementing Federalism in a Developing Country: The Case of Ethiopia, 1995-2020

This thesis assesses the implementation of federalism in a developing country by taking Ethiopia as a case. The thesis took the notion that in federations like Ethiopia, marked by developing economies, dominant party regimes, authoritarian tendencies and ethnic diversity, there are gaps between the constitution and the practice of regional autonomy. As such, the thesis systematically explores the gap between de jure and de facto by measuring regional autonomy and explaining the main drivers and instruments of de/centralisation. Regional autonomy is disaggregated into legislative, administrative and fiscal autonomy. Then, regional autonomy on two policy areas and five fiscal categories is systematically measured. The thesis reveals that the Ethiopian federation experienced centralisation over time in defiance of the constitution. The thesis shows that one of the features of Ethiopian federalism is limited regional own-source revenues. Further, centralisation of concurrent revenues and undesignated revenue powers, lack of tax autonomy, and regional lack of freedom to domestic borrowing derailed regional fiscal powers. Thus, regions are hugely dependent on federal grants. Conditional grants introduced in 2011 further eroded regional autonomy. In the spheres of policy, a range of regional policy competencies became exclusively federal in defiance of the constitution. The centralisation of land administration is a prominent example of federal encroachment on regional prerogatives.

The dominant party's structure, ideologies, and internal decision-making norms are at the core of the centralisation dynamics. The thesis shows that other factors such as regional capacity shortage, economic factors, state-building, and corruption and state capture drove centralisation over time. The thesis also reveals a de facto autonomy asymmetry across regions associated with regional development, national development priorities, peripherality, marginalisation, and ethnic diversity. Spatial and temporal de/centralisation recorded was without a formal constitutional basis through an informal constitutional amendment, proclamations, delegated legislation, organisational restructuring, and coercion. The thesis discusses that the nature of Ethiopia's second chamber and generic constitutional clauses facilitated centralisation. The thesis argues that a dominant party regime is ill-fitted for building a robust federal system as the same party controls the two tiers of government. In such a regime, though regional parties may exist, they cannot be a genuine agent of regional autonomy because of the inherent lack of independence from the federal party. In such a system, the party structure works parallel to the constitution, often circumventing the constitutional division of powers; thus, implementation of federalism is highly influenced by the party decisions. Methodologically, this implies that assessing federalism and regional autonomy in such states should go beyond the constitutional text.

Insights from Ethiopia suggest that new federations and states embracing federalism need to improve regional capacity to build effective federalism. Besides, in the long run, building a genuine multiparty system and institutions of democracy is needed for effective federalism. Federalism has unintended consequences of decentralising conflict in multi-ethnic states because elites representing different groups fight to control regional and subregional powers. Hence, new federations need to consider a mix of approaches to tackle such unintended consequences of federalism in ethnically diverse states. Lastly, moving away from a centralised legacy requires a complete political-cultural change to implement federalism effectively.