Ethiopia’s ethnofederalism: fact and fiction

While aspects of Ethiopia’s ethnofederalism may be detrimental, removing it altogether could end up causing more problems. 

9 April, 2021
by Mistir Sew

For at least a year, Ethiopian elites have been divided on whether Prime Minister Abiy Ahmed is committed to keeping, changing, or ending ethnofederalism.

For some, his rhetoric about Ethiopia’s renewed greatness and the support he gets from Ethiopianists signals a desire to end ethnofederalism. For others, his Prosperity Party (PP) remains the vanguard of ethnofederalism, much like EPRDF, its predecessor.

Meanwhile, Abiy’s public discourse and actions have not conclusively settled the matter. His Medemer book does not seem to take a strong stand on ethnofederalism’s future, and Abiy recently wrote that “Ethiopians can now imagine a future based not on ethnic chauvinism, but on unity.” Promoting “unity” is often seen as a euphemism for ending ethnofederalism.

Ethnofederalism is one way for groups to share power in diverse, divided societies. Accordingly, Abiy’s commitment to keeping or changing ethnofederalism has repercussions for groups who see themselves as staying in or being shut out of power. The Tigray People’s Liberation Front (TPLF) certainly saw itself as a shut out, which can at least partly explain events leading to the ongoing civil war.

Whether Ethiopia can successfully embark on a liberal democratic transition also depends on ethnofederalism’s future. On the one hand, elites who believe Abiy wishes to end ethnofederalism may be less willing to support or participate in democratic elections, perhaps even resorting to violence. On the other hand, if Abiy’s central government centralizes power away from the regions, it may risk negating liberal gains made during 2018.

Given that Ethiopian ethnofederalism is such a high-stake issue, it may be helpful to step back and ask “what exactly is at stake and how did we get here?” To this end, it is important to explore three questions. First, what is ethnofederalism, and how common is it around the world? Second, what causes ethnofederalism, or rather, why do those in power choose to implement it? Third, what are the consequences of ethnofederalism, particularly for outcomes such as violence, the ‘hardening’ of ethnic identities, and secession?

What is ethnofederalism?

In contrast to unitary systems, federal states are those where the central government controls some jurisdictions while the units (regions, states, provinces, republics, etc) control others. Jurisdictions may include, but are not limited to, education, taxation, or security policy. ‘Simple’ or geographical federations circumscribe units more or less arbitrarily. By contrast, ethnofederations circumscribe units to deliberately contain ethnic or linguistic groups in their historical homelands.

Scholars differ in terms of how many federal units must be ethnic homelands for a system to count as ethnofederal. For some, ‘‘at least some, if not all, the constituent units of the federation [must be] homelands controlled by their respective ethnic groups.” This describes Ethiopia’s arrangement since 1992, where some units are controlled by their respective ethnic groups (e.g., Oromia, Amhara, and Tigray) while others are not (e.g., Southern Nations, Nationalities, and People’s Region (SNNPR) and Gambella).

For others, “at least one constituent territorial governance unit [must be] intentionally associated with a specific ethnic category.” This describes Ethiopia between 1952-1962, where only the Eritrea province controlled specific jurisdictions. Whether we opt for the first or second definition determines how many countries count (and have counted) as ethnofederal.

Important current ethnofederations include India (e.g., ethnic Tamils in Tamil Nadu), Russia (e.g., ethnic Adyghes in Adygea), and Pakistan (e.g., ethnic Punjabis in Punjab), among others. Some (infamous) historical examples include the Soviet Union, Yugoslavia, and Czechoslovakia. These latter three ethnofederations broke up in a spectacular fashion and have thus received much scholarly attention. In particular, Yugoslavia continues to inform much of the popular pessimism about ethnofederalism and “Balkanization.” (Of course, it is rarely acknowledged that Yugoslavia’s ethnofederation helped manage a highly ethnically divided society for over four decades.)

This definitional point is important because some observers have tended to see Ethiopia’s ethnofederation as a unique (and perhaps uniquely evil) system. Some even associate it with apartheid. Whether or not such antipathy toward ethnofederalism is justified, it hangs on a common misconception about ethnofederalism’s commonality. To be sure, aspects of Ethiopia’s ethnofederation, such as the second parliamentary chamber’s (the House of Federation) role in adjudicating constitutional disputes, are unique.

At the same time, problems that have arisen or become more acute since Ethiopia formally adopted ethnofederalism in 1995, such as ethnic minorities’ access to jobs or land, are not unique to ethnofederalism. Indeed, ethnically-divided countries with both unitary and federal constitutions face such problems. Ethiopia’s secession clause is also unique; however, ethnofederations need not necessarily adopt such clauses, and Ethiopians could therefore decide to remove it from the constitution and remain an ethnofederation.

What causes ethnofederalism?

It is difficult to determine the beliefs and goals of elites who choose to implement specific political institutions. Elites may be under international pressure to adopt ethnofederalism (e.g., in Bosnia and Herzegovina), or they may be under domestic pressure from groups who threaten violence if their demands are not met. The latter was certainly true of the Soviet Union, where Lenin and Stalin adopted ethnofederalism against the backdrop of mobilized Georgians, Armenians, Estonians, and so on.

Whether the pressure is domestic or international, I want to emphasize the following point: ethnofederations tend to be created after unitarist alternatives are perceived to have failed. Ethnic groups may resent their status in unitary polities or in geographical federations. Such resentment can turn violent and even threaten to collapse the state.

Perceptive elites may see benefits to choosing ethnofederalism even if they find it otherwise unsatisfactory. For example, Indian Prime Minister Jawaharlal Nehru opposed ethnofederalism as a distraction from economic concerns centred around poverty and development. Yet, as aggrieved ethnic groups protested their status, Nehru gave in.

Many Bolshevik leaders saw ethnofederalism as, “not just wrong, but heretical…[ethnic] nationalism, after all, was nothing more than a clever invention of the bourgeoisie designed to deceive and divide the working classes of the world, pitting them against each other, rather than against their true oppressors.” Yet, the Bolsheviks knew they could not attempt to realize their communist ideal without endorsing ethnofederalism.

This causal point is important because it is sometimes argued that TPLF (as the founding participant in EPRDF and loudest voice in Ethiopia’s 1990s constitutional debates) chose ethnofederalism in order to “divide-and-rule” Ethiopians. The reasoning here seems to be that TPLF wanted its competitors to squabble over regional-level crumbs instead of coordinating to take the national-level pie. It is plausible that ethnofederalism serves this function. However, we must be careful not to infer the cause of some institution from the function that it serves.

Whether or not TPLF had bad intentions in choosing ethnofederalism, it is undeniable that it perceived the explosive potential of ethnic grievance in Ethiopia, much like the Bolsheviks and Nehru. There is some controversy over whether historical injustices faced by Ethiopian ethnic groups have been exaggerated for political reasons. However, it is uncontroversial that these groups believed themselves to be intensely disadvantaged under recent imperial administrations and the Derg, which were both unitarist systems. These grievances fueled several insurrections under Emperor Haile Selassie I and a civil war under the Derg. Accordingly, TPLF saw ethnofederalism as a potential solution to this costly chronic instability.

Ethnofederalism’s consequences

That recent Ethiopian history is, among other things, one of ethnic division, leads to our final question: Are the consequences of ethnofederalism such that these divisions become more pronounced or destabilizing? Although ethnofederalism may be linked to a host of outcomes—say, its relationship with democracy and development—scholars have focused primarily on consequences related to ethnic identity and violence. Although Ethiopia’s current ethnofederation has neither disintegrated nor produced any secessions (Eritrea seceded before ethnofederalism’s formal adoption), some observers worry that these outcomes are inevitable.

Critics of ethnofederalism make roughly two arguments about its consequences. First, ethnofederalism endows titular groups with control over formal institutions such as police. This enables them to more effectively secede or to commit violence against ethnic minorities with impunity. Second, the existence of ethnic homelands causes individuals to identify first and foremost as members of their group, and only secondarily with the state. This second argument also has some resonance in Amhara, where an Amhara identity was not particularly salient before the region’s creation.

But why might critics worry about ethnic identity-formation? Briefly, ethnic identities are thought to be ‘sticky’, in that individuals cannot easily change them and may view outsiders with hostility. This is less likely the case with class or ideological identities: I may be lower-class and liberal, but this can easily change. Also, it is unlikely that I view the middle-class or ideological moderates with much hostility.

Proponents of ethnofederalism concede that ethnic identities are sticky. But, they emphasize that, because ethnofederalism is usually implemented after unitarist systems have failed, such identities were likely ‘hardened’ for a long time prior. If anything, according to this second camp, ethnofederalism may be a stabilizing system as it affords groups what they have wanted all along, namely control over important jurisdictions. This even discourages them from seceding.

As an example, consider the relatively stable process by which ethnic Siltes voted to separate themselves from ethnic Gurages in 2001. Perhaps their Silte identity has now ‘hardened’, but one must compare this outcome with a hypothetical world where they did not separate, saw themselves as second-class Gurages, and possibly resorted to violence.

These paradoxes of ethnofederalism—that it seemingly makes the same outcome more and less likely—have led scholars to focus on specific background conditions. In other words, the productive way forward consists not in asking “Are ethnofederations unstable?” but rather, “Under what conditions are ethnofederations unstable?”

Broadly speaking, there are two important conditions when it comes to ethnofederal stability. The first condition concerns how ethnofederations are created. The second concerns how much the central government respects regional autonomy.

Whether ethnofederations ‘come together’ voluntarily (e.g., Canada), are created by elites in order to ‘hold together’ the country (e.g., India), or are forcibly ‘put together’ by dominant groups (e.g., the Soviet Union) has implications for stability. Unsurprisingly, ‘put-together’ ethnofederations tend to be the least stable, as ethnic groups resent their forcible incorporation and resort to violence or secession.

Although Ethiopia is usually classified as a ‘held together’ case, beliefs that TPLF forcibly ‘put together’ ethnofederalism are a source of continuing dissatisfaction. Indeed, critics bemoan the exclusion of pan-Ethiopian, Amhara, and Oromo (specifically the Oromo Liberation Front after it fell out with the TPLF in the early 1990s) interests from EPRDF’s constitutional convention.

A second important condition for ethnofederal stability is the degree to which regional autonomy is respected. If ethnofederalism is managed by dictators, it should be no surprise that ethnic groups become aggrieved. Dictators rarely respect constitutions, especially provisions that decrease their power relative to autonomous regions. In fact, observers have noted that Yugoslavia’s disintegration followed the attempts by the Serbian-dominated government to restrict regional autonomy.

Put more provocatively, Yugoslavia disintegrated not because it was ethnofederal, but because it was not ethnofederal enough. This has resonance in Ethiopia where many attribute problems of ethnic conflict to TPLF-EPRDF’s decades-long authoritarianism and unwillingness to genuinely concede autonomy to regions. And this has continued relevance as observers who worry that Abiy’s alleged commitment to unitarism will further aggrieve ethnic groups who appreciate ethnofederal autonomy.

Paths forward

Ethnofederalism is not the ideal constitutional design, but it is sometimes the most realistic option for countries with histories of ethnic marginalization and violence. Ethiopia is one such country. Marginalized groups see Ethiopia’s ethnofederation as a hard-won achievement that rectifies this history. As such, ending or radically changing ethnofederalism would provoke dissatisfaction, some of it destabilizing. This is not the path forward.

However, the above discussion points to at least two changes that would improve Ethiopian ethnofederalism. First, because ethnofederalism is more stable when regional autonomy is respected, Ethiopia’s central government should act accordingly, devolving greater fiscal and political powers to regions while refraining from intervention. On the one hand, this would appease ethnic groups that want self-rule. On the other hand, this could help demonstrate to critics of ethnofederalism the value of genuine autonomy by revealing to them that what they bash as “divide-and-rule” may actually bring them tangible benefits.

A second change follows from the fact that ethnic identities can be hard and exclusionary: elites should work to build cross-ethnic coalitions. To take a few examples: It is not only Amharas who believe TPLF engaged in an unhealthy revisionism about Ethiopian history. In addition, opposition to the current Tigray civil war from Oromo elites makes clear the potential for a possible pacifist coalition. Finally, and perhaps most pressingly, a class-based coalition would help address Ethiopia’s most intractable problem: poverty. EPRDF was initially committed to a cross-class coalition before backtracking, particularly after Meles Zenawi centralized power during the early 2000s.

Of course, it is difficult to change politicians’ incentives to intervene in regional affairs or to build coalitions. Reformers would do well to consult a recent monograph highlighting just how many options are available. These include changes to electoral rules, coalition formation rules, minority vetoes, and so on. Given the seeming increase in support for keeping ethnofederalism, this kind of incrementalism presents the best path forward.