Addressing Winner-Takes-All Politics in Ethiopia: Inclusive Majoritarianism through Liberal Consociationalism

Introduction

If one were to use a metaphor, democracy is a (football) game played once a number of ground rules have been agreed, including the shape and size of the pitch, the nature and basic organization of each team (will teams be organized [or be banned from organizing] according to height or region, or be free?), agreement on team names and insignia, manner of selection of the referee, and how awards are disbursed among winners (will all the prizes go to the winner, or will the “losers” also get a share?). Without common understanding and agreement on these fundamental issues, there would not be a sensible game.

In similar vein, democracy requires agreement on fundamental ground rules: how should the state be organized (as a unitary, federal state system or something in between, and, if federal, the manner of carving out the member states), which is often conditioned by contesting narratives, historical understandings, and interpretations (along with perceived social relations), often shrouded in ideological/normative assertions. These include questions of how political parties should or should not be organized (should identity-based parties be allowed, prohibited, encouraged, or discouraged); what may or may not be a party or state emblem; agreement on official languages; the system of government (parliamentary, presidential, or something else); how should the bureaucracy, judiciary, election management body, and other democracy, rule of law, and accountability-promoting (fourth branch) institutions be organized, etc. Without a common understanding and sufficient acceptance (or at least acquiescence) on what I consider ‘pre-democracy’ issues, free, fair, and credible democratic competition is unimaginable. Without political settlement on these issues, the result is authoritarianism all the way, and very likely an unstable authoritarianism at that.

While Ethiopians have diverged over the game of democratic elections, the core challenge arguably remains disagreement on foundational pre-democracy issues. The reality of these differences has often masked and provided catchy and comprehensible vocabulary to, intra-linguistic and inter-linguistic group contestations over power.¹ To be sure, there have been nominal claims to settling these fundamental issues. Nevertheless, these claims are often merely imposed by the idiosyncratic historical narrative and ideologies of the dominant forces of the time as the undisputed “Ethiopian” narrative.² This is partly why the 1995 Constitution of the Federal Democratic Republic of Ethiopia is sometimes referred to as the “TPLF” constitution, after the Tigray People’s Liberation Front (TPLF) that dominated the Ethiopian Peoples' Revolutionary Democratic Front (EPRDF) and midwifed the birth of the Constitution.³

Ethiopia and Ethiopians are yet to get a genuine opportunity to listen to, express, deliberate, and forge a settlement on the fundamental pre-democracy issues that must be resolved to set the stage for free, fair, and credible elections. It is in this context that civil society, opposition groups, and the broader public trumpeted calls for national dialogue, particularly following the April 2018 reshuffle within the EPRDF that catapulted Abiy Ahmed to the premiership. This was, however, not to be. Divisions and power struggles within the EPRDF, the failure of the newly minted Prosperity Party (PP) to ensure its cohesiveness and recognize, manage, and redress genuine grievances, the stubborn continuity of the tradition of authoritarian instincts, not only among the top political leadership but across the bureaucracy and law enforcement organs, some unscrupulous opposition leaders seeking to manufacture, exacerbate, and ride the wave of myopic linguistic (and increasingly religious) tensions—all these elements led to a virtual collapse of law and order and a still-ongoing full blown civil war, particularly in northern Ethiopia and parts of the Oromia region.

It is never too late to pursue the right course of action, and at the end of 2021 the Ethiopian government pushed through a legislative framework to establish a National Dialogue Commission, whose members were appointed in early 2022. The legislative framework formalized an earlier, primarily civil-society-led, initiative that set the course towards dialogue. The process of enacting the legislation establishing the Commission as well as appointment of the members of the Commission has generated justifiable rebuke and concerns surrounding the government’s commitment to the dialogue, and the reluctance of key opposition constituencies to join the national dialogue.

If Ethiopia is to break the cycle of instability, conflict, and authoritarianism and enhance its chances of charting a path towards peace and democracy, the national dialogue process would need to succeed. While national dialogue processes tend to fail more than they succeed, Ethiopia doesn’t have the luxury of squandering another opportunity. This national dialogue is critical to untie the knots that have undermined the prospects for peace in the country and sucked the energy out of democratic competition.

Ethiopia’s future is decidedly federal. In addition to addressing the many symbolic and competing historical narratives and the form and institutional architecture of self-rule, Ethiopians would have to deliberate and generate agreement on the nature and institutional manifestations of shared rule through federal institutions. This paper interrogates a modality that the ruling PP has recently unveiled to operationalize shared rule: consociational democracy—an idea that some of the ruling party’s ardent opponents have also supported.

1. Operationalizing Shared Rule: Consociational Democracy?

In addition to the broad list of symbolic and historical issues, the debate on Ethiopia’s political institutional architecture has been dominated by the nature of the federal system the country should establish. Notably, this has focused on the self-rule aspects of federalism—how the member states should be organized, what kind of powers they should have, and the means and extent of protection of minorities.⁴ An equally crucial aspect of federalism that has arguably been overshadowed and received less attention is the institutional architectures for shared rule at the federal level, which could broadly be organized along majoritarian/integrationist or consociational lines. In practice, constitutional designs often follow negotiated outcomes, combining a mixture of integrationist and consociational elements.⁵

Integrationist systems may ignore, actively seek to disincentivize, or at times outright ban political mobilization along identity lines and may incentivize ideology/policy-based politics. The political system is largely based on the individuals, who are free to organize in whatever way in their private dealings, while the public sphere formally seeks to avoid politics based on identity.

Consociational systems, in contrast, recognize and even reinforce group identity as the basis of politics. According to Arend Lijphart, the father of scholarship on consociational politics, a classic corporatist consociational system combines a parliamentary system based on a proportional electoral system, an executive “grand coalition” based on guaranteed representation of (significant or politically mobilized) identity groups, an identity-based decentralized (federal) system, and veto rights for groups on matters considered fundamental to their interests.⁶ In contrast, liberal consociational systems do not recognize or empower predetermined identity groups. Instead, liberal consociational systems entitle all political parties with an agreed-upon and notable level of electoral support to a position within the cabinet, principally in a parliamentary system.⁷ Liberal consociationalism does not necessarily require mutual group veto (as is for instance the case in Northern Ireland, which combines both corporate and liberal versions of consociationalism),⁸ nor a specific electoral system, although it arguably aligns better with a proportional electoral system. In any case, Lijphart has himself indicated that the mutual veto and proportional electoral system were secondary even to a corporate consociational arrangement.⁹

In the Ethiopian context, at the time of making of the current constitution of the Federal Democratic Republic of Ethiopia (FDRE Constitution) as well as today, it would not be an exaggeration to note that most of the political and intellectual discourse has focused on the scope of self-rule, notably on the formation of the regions (determining the “self” through language, historical connections, geography, culture, etc), the right to secession, protection of regional minorities, the status and governance of the capital Addis Ababa, and increasingly the issue of regional (special) “police” forces. This is perhaps the natural consequence of a political ideology preoccupied with the right to self-determination of linguistic groups, taken in a fundamental sense to mean the right to secession without limits.¹⁰

In contrast, the nature of shared rule has not attracted as much attention, both in political and scholarly discourse, with a few recent exceptions.¹¹ At a formal level, beyond the linguistic-based regional state organization, the FDRE Constitution adopts what could be characterized as a broadly majoritarian form of government at the federal level. Accordingly, the Constitution provides for a parliamentary system where the party or coalition of parties who wins a majority can establish a government. There is practically no enforceable specific expectation or provision as to the composition of the executive/cabinet, including, notably, being based on linguistic status. Instead, the Constitution broadly provides that linguistic groups have the right “to equitable representation in state and federal governments” (Article 39[3]). In addition, specifically in relation to the defense forces and as part of the National Policy Principles and Objectives, the Constitution enjoins that the composition of the national armed forces should reflect the “equitable representation” of linguistic groups (Article 87[1]). Perhaps one element that may make the current constitutional framework (corporate) consociational is the constitutional amendment provision whereby amendments to Chapter 3, which includes the right to self-determination including secession, require the approval of the legislative councils of all regional states, effectively granting each ethnic group with a regional state a veto over changes to the fundamental axis of the constitutional architecture (Article 105[2]).

The Constitution also provides for the representation of all recognized linguistic groups in the House of Federation (HoF) (Article 61). Nevertheless, the number of members from each group is determined largely in proportion to the size of their members, and the House is generally not involved in the regular legislative process. Accordingly, the representation of linguistic groups in the second chamber arguably doesn’t detract from the fundamentally majoritarian organization of the democratic system at the national level.

In practice, however, the structure of the EPRDF effectively supplanted the formally majoritarian democratic system in favor of equal representation of the four parties from the four biggest regions of the country within the party structure and, largely, within the cabinet and other executive entities. Members of “affiliate” parties from the other five regions had no representation within the EPRDF, although some of them had nominal representation in the cabinet—essentially forming Arend Lijphart’s “grand coalition” of representatives from the politically significant linguistic groups. This party structure was critical, as key decisions were made there and subsequently rubberstamped in the parliament. The overall objective behind the structure of the EPRDF appears to have been to empower the TPLF despite the small size of its constituency and establish what may be considered a “soft consociational” system without the potentially paralyzing consequences of formal rules or a group “veto” system. In any case, the reality within the EPRDF meant that, despite the formal presence of representatives from all members of the coalition, the TPLF, which midwifed the formation of the EPRDF on the eve of the collapse of the Derg regime, dominated the key political, security, and economic institutions, and decisions were made in closed small circles.

Prime Minister Abiy Ahmed oversaw the replacement of the EPRDF, which was the coalition of linguistic based parties, with the PP, which is formally based on individual membership, although the party still features branches for each of the major ethnic groups with their own states. The transformation was vehemently opposed by the TPLF, which refused to dissolve itself and join PP, and by Oromo ethnonationalist forces (both in the ruling party and in the opposition). These groups have criticized the party as a centralization machine and as a tool to undermine the linguistic-based political dispensation. It therefore came as a surprise when PP unveiled the “consociational system” as its political manifesto following its first party congress in March 2022. While the details remain sketchy, the establishment of linguistic/regional branches of the PP and crucially the formal reference to a consociational system signals the continued framing of politics along linguistic lines, which contrasts with criticisms of a purported official pivot by the PP away from linguistic politics, which purportedly would culminate with the replacement of the Constitution. In comparison with the ERPDF, however, a key shift in the manner of selection of the powerful central committee (which sets up the executive committee) of the PP is that, while each linguistic/regional branch nominates members to the committee, the whole party congress ultimately makes the appointments. This shift may ensure that the PP’s leadership would incentivize more moderate positions among the regional branches.

In fact, the call for a consociational system actually aligns with the views of the strongest critics of Prime Minister Abiy Ahmed and the PP. The TPLF has called for the strengthening of the linguistic-based system with guaranteed representation of ethnic groups at the national level, effectively constitutionalizing the (presumably modified) arrangement in the EPRDF. Similarly, Oromo opposition ethnonationalist leaders also seem to seek a strengthened linguistic-based politics, possibly with a consociational system at the federal level, where linguistic groups are guaranteed pre-agreed political positions.

2. Consociational Democracy: What Kind?

As noted above, the consociational model purportedly adopted by the PP is yet to be clarified. It is also not clear whether the party will seek to transform its preferred arrangement into the constitutional framework, or whether it will simply pursue it through the Party as a supra-constitutional system, much like the EPRDF did.

If the arrangement would only remain as a PP rule, it would mainly focus on a grand coalition within a single dominant party and remain a “soft consociational” arrangement, without a proportional electoral system or, crucially, a veto for any specific linguistic group. Considering that the diversified membership in the grand coalition (including officials from different groups) will come from the same party, it would be difficult to qualify it as a genuinely consociational arrangement at all. The adoption of a rule of representation in the central and executive committees of the party based on the size of population of the regional “branches” of the party represent and the size of membership may be PP’s way of operationalizing consociational democracy.

So far, it is unclear whether the ruling party will seek to elevate the arrangement to a constitutional or legal status, and may likely retain it as a party rule, which will allow it some flexibility in implementing the system. If the party were to seek to advocate for the constitutionalization of its preferred consociational system in the national dialogue, there could be two broad approaches.

The first, as indicated earlier, would involve a parliamentary system of government with a grand coalition cabinet wherein each linguistic group will get a pre-agreed share of ministerial positions, a proportional electoral system for the parliament, and veto rights for each group on matters of critical concern, alongside the (expansion of) current linguistically drawn regional states.

Beyond known criticisms of this corporatist conception of the consociational democratic model (notably its tendency to freeze and reinforce divisive identity politics)¹² the adoption of such a system in the Ethiopian context faces a number of practical challenges. First, Ethiopia is a country of minorities, and because of the sheer number of linguistic groups in the country, it would be impractical to guarantee cabinet positions for all groups, even with the assumption that these groups will each get one representative and that each is internally homogenous (which is not always accurate—for e.g., virtually all groups have individuals from different religious affiliations and of course genders).¹³ Similarly, there is a danger that granting veto powers to so many linguistic groups would lead to political paralysis and immobility.¹⁴ Indeed, consociational systems have only been implemented, and only partially, in countries with a handful of identity groups. One way around these challenges could be to guarantee cabinet positions and veto powers only to “significant” groups with populations above a threshold share of the national population. But such a solution would permanently exclude smaller groups from high positions and is thus likely to be rejected by them—they may instead prefer a soft consociational approach that at least formally leaves the route to cabinet open to everyone. Accordingly, the formal constitutional adoption of such a system may be disfavored.

If the PP is to seek the constitutional adoption of a new system, it could instead opt for what has been described as “liberal” consociationalism.¹⁵ Lijphart contrasts the genuine “self-determination” quality of this arrangement, as opposed to the “pre-determination” feature of corporatist consociationalism, and he—and a large majority of scholars—tends to favor the liberal version.¹⁶ Liberal consociationalism recognizes but de-essentializes linguistic (or religious, or any other) identity as the basis of political organization. It is a form of recognition with partial and conditional empowerment of identity (whether linguistic, religious, regional, or other) based on the inevitably dynamic support of voters.¹⁷

In liberal consociationalism, all political parties that receive more than a predetermined share of the national vote or seats in the elected federal parliament would be entitled to positions in the cabinet, making the cabinet a grand coalition of political parties, rather than identity groups. Despite its categorization as a consociational arrangement, this system effectively empowers opposition parties, whatever their color, rather than identity groups. It simply reimagines democracy in an inclusive majoritarian sense, rather than through a pure majoritarian arrangement where whichever party wins an electoral majority automatically gets to govern while the opposition represent their constituencies and focus on holding the government accountable, presenting alternative policies, and serve as governments-in-waiting. The system instead empowers and incorporates the main political groups in actual governance in proportion to their electoral support. Nevertheless, in plural societies, some major parties can be expected to organize along linguistic (or religious or regional) lines, in which case the cabinet can be expected to reflect the broader identity composition of the country, rather than simply a single identity. While it would be possible to imagine a scenario where parties that secure representation in the cabinet could be required to ensure the linguistic (or religious or regional) representativeness of their nominees to cabinet, such a rule would automatically exclude identity-based parties, undermining the neutrality of the arrangement and reducing its acceptance among ethnonationalist groups, and therefore its adoptability.

The main advantage of a system empowering opposition parties is that it transfers the power of self-determination from pre-determined and static identity groups to voters, who would have the flexibility to consider their identity (whether linguistic, religious, regional, or other) as only one factor in their voting decisions. By making the voters the deciding factors in each election round, it could be accepted as a neutral compromise among political groups, whether they favor or disfavor identity-based politics. The system can also largely be combined with a plurality (such as the first-past-the-post), proportional, or mixed electoral system, as well as with a parliamentary or semi-presidential system of government (and potentially even with a presidential system). While a proportional system could arguably encourage identity-based parties, the threshold to join the cabinet could provide a counter-balance incentive to cross-ethnic parties or coalitions. Perhaps a major challenge with this system is that it could structurally provide members of the two biggest linguistic groups—Amharas and Oromos—practical options of organizing either along linguistic or other lines, while members of smaller groups would structurally be incentivized to cooperate with other groups. This consequence may not necessarily be fatal, especially if the federal states are organized along linguistic lines, which would provide even smaller groups a platform for organizing along identity lines and articulating and defending their interests, i.e., the responsibility to protect group interests would be transferred from political parties to regional governments.

In addition to having the potential to serve as a compromise arrangement, the liberal consociational system would enable various parties to work together in the same cabinet, which could in the long term forge a cooperative and deliberative habit necessary for a thriving democratic dispensation. The presence of multiple parties in the cabinet could arguably increase the chances of government stalemate on policy issues, as has been broadly argued in relation to consociational systems. Nevertheless, if the national dialogue process helps resolve the key symbolic and institutional contestations underlying political and security disputes, differences on ordinary political, social, and economic issues are unlikely to evade compromise. In any case, the main party or coalition of parties can be expected to cobble together sufficient majorities within the cabinet whenever broad consensus proves elusive, as liberal consociational arrangements do not necessarily grant veto rights to any group or party.¹⁸ Nevertheless, even when decisions are taken through a vote, losing parties and their supporters can be sure that their voices have been heard at the highest level, which would increase the chances of consent among losers.¹⁹ Indeed, one key advantage of liberal consociationalism is that at least some of the opposition parties would win positions, which reduces the winner-takes-all mentality, and therefore gives such parties a stake in recognizing electoral outcomes, thereby reducing the possibilities of violence.

If the PP would maintain consociationalism as a working arrangement within the party rather than seeking to constitutionalize or otherwise legalize the arrangement, it is possible that the current, largely majoritarian, arrangement for shared rule could continue at the formal level. The current arrangement has the advantage of leading to a relatively coherent central government, which a government formed based on a liberal consociational arrangement may not always deliver. In such a case, the composition of the cabinet and other aspects would be open to pre- and post-election political negotiations. The inclusion of opposition parties in the cabinet will also largely fall to the discretion of the ruling party or coalition.

Whether or not a consociational liberal system is constitutionalized, Ethiopia should consider strengthening the position of opposition parties to not only effectively serve their representational, accountability providing, and government-in-waiting roles, but also include them in governance. This could, for instance, take the form of public funding to opposition parties, guaranteed equitable access to publicly-funded media throughout the year (rather than merely during election periods), and, crucially, the constitutional recognition of a Leader of the Opposition, alongside publicly provided resources and secretariat, as well as security and diplomatic privileges. Constitutions around the world have adopted creative and systematic ways of promoting what may be considered “inclusive” majoritarianism that eschews winner-takes-all politics, thereby enhancing the vitality and resilience of democratic systems.²⁰ Ethiopia could consider a systematic constitutional and political arrangement combining both liberal consociationalism and broader ways of operationalizing inclusive majoritarianism.

Conclusion

This short contribution has sought to shift attention that is often focused on self-rule aspects of Ethiopia’s political dispensation to equally important arrangements for shared rule. Whatever the fate of the national dialogue process, in view of the political dynamics, the organization of political parties, and the inertia of the status quo, the federal arrangement with regional states drawn along linguistic lines can be expected to stay and perhaps be reinforced with the formation of additional states. To be sure, even if the linguistic-based regions remain, the national dialogue could still be considered successful if it helps to build an acceptable political settlement among a critical mass of Ethiopians around: a narrative of the country’s past and symbolic issues (e.g.. the flag); the identity of the nation (including potentially by recognizing both Ethiopians as a whole and each linguistic group as a “people”);²¹ the fundamentalism regarding the right to self-determination (especially the right to secession); the recognition, rights, and systematic and effective representation of all groups (notably political and linguistic minorities) at the regional level (potentially through the application of liberal consociationalism at the regional level); and around the status and governance of Addis Ababa. Meaningful compromises on these could make the linguistic based federal structure palatable even for those who vehemently reject it.

These issues related to self-determination have received significant attention in the political and scholarly discourse from Ethiopians and non-Ethiopians alike. This short contribution has sought to highlight an important area—shared rule and the form of democracy at the national level—that is necessary for a complete understanding of the full universe of issues that Ethiopians need to grapple with to move towards a sustainable peace and inclusive democracy. To be sure, Semir Yusuf has developed an excellent analysis and outlined potential options—without proposing a particular model—in designing both self- and shared-rule aspects of federalism in a diverse polity by combining elements of consociational, centripetal, and integrationist ideas to manage pervasive ethnic divisions.²² In another contribution, the author of this paper (Adem Kassie Abebe) has outlined potential ideas on how to structure Addis Ababa in an inclusive and autonomous manner, largely drawing on ideas from these divergent concepts of liberal consociationalism, and has suggested that experience with such an arrangement in the governance of Addis Ababa could provide useful insights for reevaluating the thinking and design options at broader national and regional state levels.²³

In essence, the core point is that, despite the recent announcement that the PP has adopted consociational democracy as its organizing principle, this is likely to remain a party rule—with little prospect of elevation to a constitutional/legal rule. Even if it becomes a constitutional or legal rule, it would likely take the form of a softer version of inclusion rather than Lijphart’s sense of a formula-based grand coalition and mutual veto. This is primarily because the system is impractical in view of the sheer number of identity groups in Ethiopia. It is conceivable that the biggest groups could be granted a sort of soft share of power in the decision-making bodies of the Party, but even such a flexible arrangement is likely to be left open regarding the cabinet and other key state institutions. In any case, a consociational arrangement within the ruling party is unlikely to address demands for inclusion, as some groups would see purported representatives from their group as ineffective and even nominal. Despite the essentialization of linguistic identity among ethnonationalist forces, intra-group contests have often led to outbidding efforts and denial of the linguistic identity of rivals, which appears to run counter to the essentialism that seeks to portray a unified and cohesive identity.

A more likely approach is to pursue reforms towards the inclusion of the opposition in governance, including through liberal consociationalism as outlined in this brief paper.

The central attraction of the proposal, and potentially key to its acceptability among contending forces, lies in its dynamism and avoidance of pre-determination, along with the empowerment of the people to entrust power to whichever political ideologies and identities they may prefer in each election cycle. Furthermore, the proposed suggestions can work in both parliamentary and presidential systems of government, or in any other system. The arrangement would partly eschew controversies around the population size of linguistic groups, as the proposed system would ensure representation based on popular support in each election cycle rather than the population size of any specific group. Moreover, while the proposals are mainly focused on the national level, the ideas can be equally useful to consider in organizing regional states. Indeed, the author of this paper first proposed these ideas for the governance arrangement for Addis Ababa.

This paper has mainly focused on the federal executive and principal legislature. There are of course other platforms for the direct representation of linguistic and other identity groups. One option is the second chamber, as Ethiopia currently has. Secondly, regional and other lower-level governments are perhaps more appropriate fits to represent the interest of groups than national representatives elected on partisan basis. But this would require a channel of communication where the views of regional and other levels of government are sought out and allowed to feed into the conversations in national legislative and executive decision-making processes. Intergovernmental deliberation platforms, involving not only regional but also local governments, are therefore crucial, as is the manner of representation emanating from liberal consociational arrangements considered in this paper.

Discussions

Abdulatif Khedir — Discussant

In your paper you consistently used the term “linguistic groups” as opposed to the commonly used term “ethnic groups;” are you being a liberator on this and is it intentional? Because I would assume that some may perceive describing segments in the divided Ethiopian society as simply linguistic as a bit reductive. So, if you are using the term “linguistic group” deliberately I think you need to say a bit on that.

The other point is, from the different segments of society, maybe because it is the most salient, there is an exclusive focus on what you call “linguistic groups” and what others call “ethnic groups,” but what about other segments such as religion? Of course, there have to be people mobilized along this line but can you take the current mobilizations, antagonisms, and conflicts along religious lines as an analytical angle, dictating our desired options when discussing consociationalism and consociational power sharing?

Another point is that you focus on the shared-rule aspect of consociationalism, but segmental autonomy is also a primary feature of consociationalism; I think you need to give it more space in the paper. This is important because segmental autonomy in consociational sense can be implemented in a non-territorial way. So far, our country’s political discourse is all about autonomy for territorially organized groups, but there are a lot of issues that cannot be addressed simply by territorial arrangement. Take for example the case of groups who are not territorially settled in one area or whom we call dispersed minorities; they may enjoy segmental autonomy in the form of legal pluralism or devolving certain affairs to communities, and this can be important to deal with non-territorial autonomy issues.

Another focus of your paper is power sharing or grand coalition as one basic feature, mostly in the parliamentary sense. But consociationalism can also be implemented in a presidential sense. And a lot of political parties and groups think that the presidential system might be good; there is good chance that the future in Ethiopia is a presidential system. Given this prospect, it may be helpful if you reflect on what consociational elements would be appropriate in a future presidential arrangement in Ethiopia.

Another point you raised is the difficulty of organizing grand coalitions given the sheer number of linguistic or ethnic groups, with other potential dimensions. One interesting suggestion I remember from the late Professor Mesfin entails guaranteed seats in cabinet for groups that number a million and above and, for the others, having cabinet seats through rotation. It is good to reflect on this.

We are so far talking about consociationalism at the national level. But it can also be very important at the subnational level. Whether the current federal structure stays intact or is redrawn along territorial lines, or even further along ethnic lines, we will have permanent minorities, especially at the regional level. It is important to consider how consociationalism can be implemented at the subnational level to cater to the needs of these minorities. Further, if the consociational arrangement is tried and succeeds at the regional level, I think it can be a good lesson to implement it at the federal level. There are also a few examples. I think Harari as well as Dire Dawa (the rotation of the mayoral position between the Somali and Oromos) are interesting to closely look at; although they are not strictly consociational arrangements, they have elements or features of it.

You also raise the issue of proportionality, not only in the representation in the grand coalition that is in the executive but also in the civil service and all other organs of the State, including the judiciary. But there is a challenge in numbers. Numbers are important in proportional representation arrangements, because when you are talking about proportionality you are talking about the allocation of seats and benefits depending on the number of people that some particular political group identifies itself with. But population numbers in this country are increasingly contested. Each ethnic group claims to be huge in number, so much so that it will make Ethiopia a country of around 300 million people if we accept all the claims made by the elites of the various ethnic groups as to their population size. Given this contest about numbers, the implementation of proportional power sharing arrangement would be problematic.

Finally, one of the reasons why I like the idea of consociational power sharing at the center is, even if there is a centralizing tendency as emphasized by other presenters, consociational power sharing will help to ensure the legitimacy of the center.

Dr. Zelalem M. Teferra

When Adem started his presentation I wished he would close his presentation with a quote from an article by Dr. Dereje Alemayehu titled “ጀግና የማያስፈልጋት ሃገር የታደለች ነች (ǧagenā yamāyāśefalegāte hāgare yatādalače nače /Lucky is the nation that stands in no need of a hero (a saying attributed to Bertolt Brecht)” in which he states that Ethiopia will never be peaceful until “በቁጥር መበላለጥ የመብት መበላለጥ የማይኖርበት አገር አስካለፈጠርን ድረስ (baquṭere mabalālaṭe yamabete mabalālaṭe yamāyenorebate hāgare ʼeśekālefaṭarene deraśe/ we build a nation where population size [of political groups] doesn’t make difference for protection of rights.” I think it is good to keep this in mind while discussing consociational governance.

Abdulatif Kedir

Additional point on Adem’s presentation: In relation to centralization, I feel that there is a deserved critical appraisal of how the centralization process was conducted and how it resulted in the suffering or problems of the country. But I think there are at least two reasons why we also need a legitimately stronger central government. One issue that we have overlooked is the issue of economic development; poverty in this country is the real problem and we need a strong and committed government at the center to address this problem without neglecting the subnational units. The other is the issue of minority protection. Some suggest confederal arrangement to this end, but I do not see an alternative beyond a strong but legitimate and representative central government as a guarantor of minority rights protections in the sub-national units. Some proponents of the current federal arrangement suggest that further decentralization along ethnic lines will solve the problem, which I do not think is the case. Others think redrawing the federal units will solve the problem. In any event we will have subnational minorities and I think a strong center is a guarantor of their protection because I think most sufferings and violations of rights in recent years are attributed to weak central government.

Dr. Mulugeta Mengist interjected and said: instead of “strong government” I would say “effective central government;” when we say strong government it tends to be forceful, but being forceful does not translate into effectiveness.

Dr. Getachew Assefa Woldemariam

When you talked about inclusion and empowering the opposition, you said that the effective way of doing it is to include all regional and subregional government institutions. The question I have is: Do you make an assumption that governments from regions are politically or representationally diverse and well-representative, or are you talking about the kinds found in the current arrangement? If you are talking about in the latter sense, given the similarity of the occupants of the various assemblies, would that make any difference? Are you assuming a recasting of the current arrangement in favor of democratic diversity in terms of political representation for it to work? Other than that, how can your suggestion that there must be a government that solves society’s problems work when there is no representative government?

Dr. Abadir M. Ibrahim

I want to raise an important point to you all. One thing we have not talked about, and this is something I constantly wonder about, is the importance of local democracy, city democracy and so on. I understand the importance of the center. But I have the sense that democracy really happens at the local level, and it trickles up to the center rather than trickling down from the center. Are we ready for pluri-experimentation of different types of local governments? Usually solutions come from the center, i.e., policies are designed at the center and go to the local level to be implemented. Even if it comes from the center, are we ready for differentiated experimentation? We have to try a bottom-up approach to democratizing the state.

The other point I want to emphasize is the participatory, big-tent kind of political process from the point of view of transitional justice. It is important not only in the long run, but also in the short run in the process of state formation, which is itself the outcome of a transitional process. There are always going to be losers and winners in constitutional design but when you have significant actors in society that become losers, that means you will have even more permanent losers going beyond minorities (with non-minorities becoming losers) because they feel that they did not, or in fact did not, participate in the constitution-making process, or their interests were not reflected sufficiently in the process.

Notes