Ethiopia's Experiment in Accommodating Diversity: 20 Years’ Balance Sheet

Abstract

After years of centralised rule that emphasised unity, since 1991 Ethiopia has adopted a federal system that aims at accommodating diversity. This article assesses whether Ethiopia's experiment with federalism is achieving the promised goal of managing diversity or whether it is exacerbating conflicts. The federation's biggest risk is its entanglement with the ruling party, its reliance on soft institutions of democracy and narrowly based institutions of power sharing. Although states are given broad mandates to design policies that fit their local context, in actual terms, their role is limited to implementation of policies designed at the centre.

Keywords:

• Ethnic Federalism Ethiopia
• Accommodating Diversity
• Minority Rights
• Federalism and Democratization

Acknowledgements

The author would like to thank two anonymous reviewers for their important remarks.

Notes

Among the PhD dissertations, see Solomon (2006); I make no attempt to address fiscal issues here; Assefa (2006a), Yonatan F. (2008), Van der Beken (in press) and Asnake (2009).

Regrettably, as far as the case studies from Ethiopia are concerned, the conference was more of a public relations exercise, notable experts of Ethiopian federalism either missing or playing little role.

See, for example, Minase (1996: 1–84) and Brietzke (1995: 19–38). A notable exception is Abbink (2006; 2009), who has discussed extensively the challenges of democratisation and the nature of political culture in Ethiopia—very crucial points in our understanding of federalism in Ethiopia, considered towards the end of this essay.

On the contrary, Kymlicka (2007: 135–136) rightly demonstrated that

evidence of the last forty years of multiculturalism in the west shows that those that adopted it never descended to civil war, anarchy or faced military coups or suffered economic collapse. On the contrary they are peaceful, stable and prosperous. This refutes some of the arguments about the destabilizing effects of multiculturalism, that is as the first step on a slippery slope towards ethnic strife, balkanization and local tyranny.

The secession clause is referred to in Article 39 of the 1995 Ethiopian Constitution and it is one of the most contested provisions of the constitution, about which a lot has been written. Article 39 though is a broad one and is not limited to secession rights alone. It provides for the right to self-government of ethno-nationalist groups, including fair representation at all levels of governments and the procedures of secession. Its inclusion in the constitution, apart from the fact that it was a long-held view of the current ruling party owing to its leftist inclination, was justified as a means of bringing the national liberation fronts, such as the Oromo and the Ogaden National Liberation Fronts, to the negotiation table because of the agenda of secession. Yet the experience of Eritrea, (Eritrea seceded from Ethiopia in 1993 and is currently neither internally peaceful nor peaceful with its neighbours) and other new states hints that secession is not the right means to address a crisis of governance. It replicates the same old normative political issues in the new state. Hence experts suggest more accommodative political solutions, such as federalism, power sharing and decentralisation, as a way out. See Horowitz (2003).

Federalism, as defined by the late Daniel Elazar (1987: 12), is a combination of “shared rule and self-rule”. A more elaborate definition would include five elements: (1) two levels of government (one for the whole federation and the states for the units) rule the same land and people; (2) each level has some area of action in which it is autonomous, guaranteed by the constitution and not subject to unilateral change; (3) the states are represented in the federal policy-making usually, but not exclusively, in the second chamber; (4) the polity provides some form of interaction, formal or informal, for joint action of both levels of government on common issues; and, finally, (5) it has institution for mediating intergovernmental conflicts (Watts, 2008: 9).

The phrase ethno-nationalist groups refers to groups found regionally concentrated and that conceive of themselves as nations within a larger state and are mobilised behind an ethno-nationalist party to achieve representation/recognition of their nationhood either in the form of independent states or through territorial autonomy within a larger federation. Compared to other forms of social mobilisation, such as class and gender, there is a visible saliency of ethno-nationalist mobilisation by such groups (Kymlicka, 2007: 68; Hale, 2008: 2–3).

15 million (20% of the total population) of which are found in the Southern regional state, an amalgam of some 56 ethnic groups. The total population estimate of 2007 was 74 million. See Summary and Statistical Report of the 2007 Population and Housing census results, issued by the Central Statistical Agency (CSA), available online at http://www.csa.gov.et/pdf/Cen2007_firstdraft.pdf (accessed 29 August 2011).

The Ethiopian state, in its loose sense, begins with Axum (first century AD to approximately 1150; Levine, 2000: 90–110; Bahru, 2001).

The exceptions include Yodit/Gudit's attack on the Christian empire in the tenth century; the campaign of Imam Ahmed (1527–43) and the Age of Princes (Zemene Mesafint, 1769–1855). The power at the centre had to subject itself almost completely to regional forces.

Ethiopian legend has it that Menelik I, son of the Queen of Sheba of Ethiopia and of King Solomon of Israel is the founder of the Ethiopian state that was based in Axum and all other Emperors had to claim this legend to legitimatise their power (see Markakis, 1974: 43–71).

The meaning of this notion is far from clear. Young university students in the 1960s and 1970s, influenced by Marxism-Leninism and frustrated by the age-old monarchy, argued in favour of liberating ‘oppressed nationalities’ from an oppressor group but, as later events indicate, it meant different things to different groups. For some it meant respect for diversity but to be implemented under a class perspective. For the now ruling party, Ethiopia is nothing but the outcome of the free will of the nationalities rather than abstract individuals and the nationalities have not only the right to self-rule but also the right to self-determination, secession included. See Articles 8, 39, 46, 47 of the 1995 Ethiopian Constitution. For details, see Young (1997).

Ethiopian People's Revolutionary Democratic Front (EPRDF), established in 1989, is a coalition of four ethnic-based organisations—the core and founder, Tigray People's Liberation Front (TPLF), the Amhara National Democratic Movement (ANDM), the Oromo People's Democratic Organization (OPDO) and the Southern Ethiopian People's Democratic Movement (SEPDM). It came to power in 1991 and is still the ruling party today (Vaughan, 1994).

Nearly all centrist and mainstream Ethiopian writers fall into this category (Bahru, 1991: 60–71; see also Teklestadik, 1982: 296; Maimire, 2006).

Such thinkers are called instrumentalists because they view religious- and ethnic-based mobilisation by groups as a means to achieve some political and economic goal and having no relevance thereafter. Included in this category are Markakis (1998), Clapham (1994: 30–35) and Abbink (1998: 60).

For details on the failure of the nation-state, see Young (1974: 507).

This reminds us of the prominent article by Connor (1972), where he argued that modernisation, industrialisation, respect for individual rights do not necessarily minimise ethnic loyalty, and nation-building may in the end be nation-destroying in diverse societies.

This expression refers to cases in which identity-based politics have a high degree of prevalence exceeding that accorded to alternative forms of political mobilisation, such as ideological, class, gender, and the relationship between groups is affected by deep levels of mistrust making it less co-operative (see Horowitz, 2002: 18).

Is it political power, resources, the military that are to be shared? See, for example, Andreas Mehler (2009: 2–10).

According to Richard Simeon (2008: 55), the diversities are diverse in themselves owing to their historical origin, manifestation and type of diversity, numerical size, being territorially grouped or dispersed, difference in level of mobilisation and the gravity of repression committed by the state.

The exceptions are Benishangul Gumuz and Gambela and partly the Southern Nations, Nationalities and Peoples regional state (SNNPRS) (see below).

The SNNPRS is peculiar in some respects, with an estimated population of 15 million (20% of the total population). This is a constituent unit with extraordinary heterogeneity and relatively smaller ethno-linguistic groups in which local governments (in the form of zones, weredas or special weredas—sub-state-level local governments), not regional state governments, as in the six other states, are designed as a means of ensuring self-rule to the various groups. There is also the effort to share power and resources among the various groups in the regional state institutions. It is a kind of federation within a federation (Vaughan, 2006; Assefa, 2002; and Habib, 2010: 155–156).

The same explanation has also been stated by Horowitz (1985: 131–134).

The preamble of the Ethiopian constitution stipulates “Fully cognizant that our common destiny can best be served by rectifying historically unjust relationships …”.

Examples include Spain (Catalans and Basques), Canada (Quebec), Switzerland (French and Italian minorities), Scots and Welsh in the UK, and partly India.

The risks are well known in the literature on comparative federalism. The existential secession threat by weakening institutions of the centre in favour of the units; that local-level majorities may turn themselves into autocrats and terrorise minorities under them; frequent claims for border adjustments and various forms of local self-rule issues (Nigeria and SNNP in Ethiopia); emphasis on the politics of difference over cross-cutting values by ethnic entrepreneurs that may then aggravate conflicts. For example, ‘two solitudes in Canada; three groups with their backs to each other in Switzerland, hard and open ethnic identities in Ethiopia (see Brown, 2007: 57–78; Roeder, 2009: 204–208).

The terminology was coined by Alfred Stepan (1999).

For a more comprehensive treatment on the division of legislative and executive power between the federal and state governments in Ethiopia, see Fiseha (2006b: 131–164).

See, for example, the Indian Constitution, Articles 258(1), 154.

Article 50(9) and for the record of the debates, see Ye Ethiopia Hige Mengist Gubae Kale Gubae [Minutes of the Constitutional Assembly], Vol.4 Hidar 14–20, 1987 E.C. (November 1994) discussion on Art. 50.

See the Minutes mentioned above.

According to Article 52/2d, regional states are mandated to “administer land and other natural resources in accordance with federal laws”. Based on this clause, regional states have been providing land for lease to investors in their jurisdiction. Allegedly, the federal government claims that regional states have not been efficient in providing land to investors and complained about widespread corrupt practice across regional states and, hence, the need for upward delegation. Interestingly, such widespread corrupt practice is also common at the federal level, as openly presented to the public through government-owned Ethiopian television on various occasions. Thus, some doubt the plausibility of the federal government's argument.

The HoF also settled the dispute between indigenous ethnic groups and non-indigenous groups in Beni-shangul Gumuz; see below the section on the position of minorities in regional states.

See Articles 72, 73 and 74.

See Articles 70 and 71.

Asymmetry in federal state relations implies that either the federal government enjoys more powers in some regional states than others or some of the regional states enjoy more powers than other regional states, as in the case of Quebec in Canada or Catalonia in Spain (see Watts, 2008: 125–127).

The two-tier nature of the federal system (those with relatively better experience in self-rule versus marginalised ones) has been made clear in a number of studies. See, for instance, Young (1999). Abbink (1997: 173) also remarked that in the constituent states under discussion, there have been dismal failures.

For details, see Ethiopia ye 2001 Ametawi Metshaf (in Amharic) [Ethiopia 2001 E.C. Annual Book] (Addis Ababa: Master Printing House).

A major dividing issue in this respect is that the increase in education at all levels, without the necessary preparation in terms of qualified personnel, libraries and other necessary inputs, has compromised quality significantly.

In addition to the competition over resources between the Afar and Issa, which is more or less the same case as with the Borona and the Gari, the tension between the Afar and Issa becomes more complex because it is very much interlinked with the geopolitics of the Horn. Situations in Djibouti, now Ethiopia's only access to the sea, in which the Issa have the political upper hand, and the minority position of the Afar in Djibouti, and intervention by neighbouring Somalia and Eritrea, complicate the traditional conflict between the two ethnic groups. For more on this, see Markakis (2003).

For instance, several separatist parties backed by neighbouring Somalia interfere in the Somali region and the Oromo Liberation Front does a similar job in other areas.

Article 48(1). The HoF is expected to render a final decision within a period of two years. Often it tries to create a forum for negotiation between the states but if that fails it facilitates a referendum. See Article 62.

This referendum was conducted in 2008. I am grateful to Tekle Didu who brought this case study to my attention while he was working on his MA thesis in this area in 2009.

Non-Oromo here implies either smaller ethnic groups that exist within Oromia regional state or individuals from ethnic groups other than Oromos. For details, see Zerihun (2009).

See proclamation numbers 65/2003 of the urban local government of the Oromia regional state Megeleta Oromia 9^th year No. 2 Adama 2003 and 116/2006 to amend the former Megeleta Oromia 14^th year No. 12 Finfine 2006.

For details, see Sisay (2009).

For example, the Kunama minority in Tigray are found in two districts of two different weredas dominated by the Tigrayans. The Irob have their own wereda but have not yet started using their Saho language.

According to the 2007 census, the population of the Harari regional state is 183 344, of whom 103 421 are Oromos and 41 755 are Amhara.

See Harari State Constitution Arts 26 and 29. Only the Hararies and Oromos can participate in the region's politics.

At the core of a Lijphartian type of consociational democracy is that every relevant group shares power and resources in proportion to the size of the vote and hence there is no exclusion of any group whatsoever.

For the historic-political background of both regions, see James (2002).

According to the CSA census of 2007, Gambella is home to five indigenous ethnic groups—Anuak 21.2%, Nuer 46.6%, Majengir 4%, Opo and Komo 0.4%; Benishangul Gumuz is equally home to Berta 26%, Gumuz 21%, Shinasha 7.6%, Komo 0.96% and Mao 1.9%.

In Gambella the highlanders constitute about 27% while in Benishangul Gumuz the same category of people constitute nearly half the population of the region.

The decision ensured the right of non-indigenous minorities to be elected to public office as they were excluded by local parties from the process. See the decision of the HoF of Megabit 5, 1995 EC (March 2003), which basically upheld the then existing law (Proclamation 111/1995) as constitutional but declared the decision of the Electoral Board that excluded the non-indigenous groups from running for office as unconstitutional and, hence, of no effect with prospective effect. It articulated that Art. 38 of the proclamation is not in violation of Art. 38 of the Constitution. It also underscored that whoever wants to run as a candidate is required to know the working language of the regional state, and not one of the local vernaculars (unpublished).

For a detailed account of the human rights crisis, see Human Rights Watch (2005).

The book was first published in 1974 and has now been reprinted again.

According to Erk (2007: 262–263), the study of federalism can be evaluated by four benchmarks: the links between federalism and democracy; federalism's role in accommodating territorially grouped diversity; federalism's impact on policy-making; and, lastly, the design of federal institutions in responding to the three preceding issues.

The GTP is based on an estimated annual economic growth of 11–15%, achieving the Millennium Development Goals, and transforming Ethiopia to the category of middle income countries by 2020–23, among others. For details, see http://www.mofed.gov.et/English/Resources/Documents/GTP%20English2.pdf (accessed 20 August 2011).

While many of the ethnic-based opposition parties agree on the federal option, they criticise the ruling party for failing to deliver as promised in the Constitution.

Nigerian federalism in this period was a parliamentary one where leaders of the regional parties chose to head their respective regional governments and send deputies to the federal level, thus resulting in weak federal government and strong states and finally resulting in the attempted Biafran secession of 1967–70. The regional political leaders, for example, Chief Awolowo, was Premier of the West, Dr Nnamdi Azikiwe was Premier of the Eastern region and Sir Ahmadu Bello remained Premier of the North (see Ayoade, 1996: 52; Barkan, 2004; Mueller, 2008).

Some have already hinted that “while federalism may survive the EPRDF, the present political framework is so dependent on the regime that created it and its survival is unlikely” (Clapham, 2009: 191).

Abbink (2009: 4–5) argued, “it is a political model that is risky …”; ethnic-based federalism does not work and produces ethnic-based local conflicts.

Failed former communist federations are examples.

On what constitutes the opposition in Ethiopia, see Asnake (2011).

According to Abbink (2009: 12), “in the post 2005 traditional authoritarian culture was confirmed closing the space for republican democracy”. See also the European Union report on the 4^th national and regional state council elections (European Union Election Observation Mission, 2010), which concluded that seen from two angles—the lack of a level playing field for all contesting parties and the narrowing of political space—it failed to meet international standards.

The National Election Board stated that the ruling party and its allies secured 347 seats in parliament, with a seat total of 547, and declared it winner. The CUD secured 109, while the United Ethiopian Democratic Forces (UEDF) won 52 seats. Eleven seats went to an Oromo opposition, with one going to an independent candidate. See NEBE announces official results at http://electionsethiopia.org/Whats%20new40.htm (accessed 15 June 2006). A majority of the members of the CUD boycotted the new legislature's first sitting, but 93 of them later joined the HoPR. Members of the UEDF, the second largest opposition, and other smaller opposition parties, however, joined Parliament from the outset.

In Addis Ababa, the opposition CUD won 137 (99.3%) in the 138-seat parliament. By some twist of history, the April 2008 by-elections resulted in 137 (99.3%) seats going to the ruling party and only one seat going to the opposition.

See comments by Christopher Clapham on the Ethiopian crisis written in mid-Nov. 2005, available at http://www.ethiomedia.com/fastpress/Clapham_on_Ethiopia_crisis.html.

The joke in town is that such organisations are divided into two: teachers’ association number one belongs to the ruling party and teachers association number two belongs to the opposition. There are two Bar Associations and one is allegedly affiliated to the ruling party and the other to the opposition. All other forms of civil society groups (except those traditional self-help ones and a few philanthropic ones), including the youth, women and farmers, are branded in the same way and this perception proves the bulk of them are not genuine and autonomous civil societies.

“The Ethiopian government was more narrowly based and had less legitimacy in 1994 than in 1991. The major opposition movements had moved outside the legal political process” (Ottaway, 1995: 73).