Consociationalism for Weaklings, Autocracy for Muscle Men? Determinants of Constitutional Reform in Divided Societies

Abstract

Why have some countries adopted consociational constitutions after civil wars, while others have not? This contribution analyses the advent of constitutional provisions and their rationale in the five deeply divided societies that adopted the most ‘radical’ constitutional reforms in 2001–10 (Burundi, Comoros, the Democratic Republic of the Congo, Nepal and Sudan). In places where norm-diffusing external actors played a major role in ending an intra-state war, constitutional change was particularly profound. It appears that strong outside mediators have followed a blueprint approach to constitutional engineering in post-war societies with power sharing as a guiding principle, more often than not of the consociational type. There is reason to believe that the strong outside involvement in turn depends on a relative weak bargaining position of sitting governments.

Notes

^1. I consider ‘radical’ a constitutional reform that changes main regulations of most if not all main formal institutional areas (e.g. security sector, judiciary, party system, elections, state and government system).

^2. Helga Binningsbø, ‘Power Sharing, Peace and Democracy: Any Obvious Relationships?’, International Area Studies Review 16/1 (2013) pp.89–112, p.89 calls it ‘a dominant conflict-solving approach’.

^3. Caroline Hartzell and Matthew Hoddie, Crafting Peace. Power-Sharing and the Negotiated Settlement of Civil Wars (Pennsylvania: Pennsylvania State University Press 2007), pp.43–63.

^4. Hoddie and Hartzell also test the influence of the impact of the international system (i.e. during or after Cold War), this is not a distinguishing factor in the cases under scrutiny here (all civil wars terminated after the Cold War).

^5. Arend Lijphart, Democracy in Plural Societies (New Haven, CT: Yale UP 1977) p.25 et seq.

^6. E.g. Donald Horowitz, Ethnic Groups in Conflict (Berkeley, CA: UCP 1985, 2000) or Ben Reilly, Democracy in Divided Societies: Electoral Engineering for Conflict Management (Cambridge: CUP 2001).

^7. Horowitz (note 6); Timothy Sisk, Power Sharing and International Mediation in Ethnic Conflicts (Washington, DC: USIP Press and Carnegie Commission on Preventing Deadly Conflict 1996); Reilly (note 6).

^8. Andreas Mehler, ‘Peace through Constitutional Amendment? Opportunities and Tendencies’, GIGA Focus International Edition 2 (2011) pp.1–8.

^9. I have opted not to include Europe, the Pacific, North America and the independent states of the former Soviet Union due to my limited research capacities. Two cases may be lost by this deliberate decision: Macedonia which had a brief violent ethnic conflict, a peace agreement and subsequently a change in Constitution in 2001 taking into account the interests of national minorities, and Bougainville/Papua New Guinea with equally a comprehensive peace agreement in 2001 and a new Constitution in 2004. I have also not included the prominent cases of externally mediated elite pacts that ended post-electoral violence in Kenya, Madagascar and Zimbabwe. Those are not considered civil wars and do not end up in the relevant data-sets. Let me just note that for the constitutional reform processes in all three countries, those crises and their resolution had crucial importance.

^10. The Minorities at Risk (MAR) Dataset. College Park, MD: Center for International Development and Conflict Management. The MAR programme has identified 283 groups with at least 500,000 members (a very arbitrary choice as the size of the polity is not taken into account). Not all groups are evident and well defined (e.g. ‘Westerners’ in Cameroon, but no ‘Northerners’ in Côte d'Ivoire). Twenty-nine groups are in the MENA countries, 75 in Africa, 57 in Asia and 33 in Latin America. Neither Nepal nor Côte d'Ivoire is in the database, both can be considered as important gaps. Other cases could be discussed: Yemen, Comoros, etc., online at < http://www.cidcm.umd.edu/mar/data.asp>, accessed 8 Mar. 2013.

^11. Lars-Erik Cederman, Brian Min, and Andreas Wimmer, 2009-05-01, ‘Ethnic Power Relations Dataset’, online at http://www.hdl.handle.net/1902.1/11796, accessed 8 Mar. 2013. See Andreas Wimmer, Lars-Erik Cederman, and Brian Min, ‘Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Dataset’ American Sociological Review 74/2 (2009): pp.316–37.

^12. Lars-Erik Cederman, Brian Min, and Andreas Wimmer, ‘Ethnic Armed Conflict Dataset’, online at < http://www.hdl.handle.net/1902.1/11797> V1, accessed 8 Mar. 2013. The EAC database codes armed conflicts as ethnic if they are conflicts over ethnonational self-determination, the ethnic balance of power in government, ethnoregional autonomy, ethnic and racial discrimination, and language and other cultural rights.

^14. I had to modify the UCDP coding quite strongly as peace agreements or ceasefires in intra-state wars in Burundi (2002, 2003, 2009), CAR (2008, 2009), Chad (2002, 2005, 2006, 2007), Comoros (2001), DRC (2003), Mali (2009) and Niger (2009) were not included in the database. Uganda may be a limit case: I have coded the 2006 ceasefire between government and LRA rebels as important enough to be included, this may be debatable. I have followed the advice of my colleague Sandra Destradi to term a whole number of ceasefire agreements in India's local conflict during the observation period as not important. For peace accords in Africa 1999–2007 see Andreas Mehler, ‘Peace and Power sharing in Africa: a not so obvious relationship’, African Affairs 108/432 (2009) pp.453–73.

^15. In an earlier paper (Mehler, see note 8), I had included Iraq because a ceasefire agreement with the Al-Mahdi army was enacted in 2007/08, but this looks in retrospect of marginal importance (ended in March 2008 by a government army offensive).

^13. Joakim Kreutz, ‘How and When Armed Conflicts End: Introducing the UCDP Conflict Termination Dataset’, Journal of Peace Research 47/2 (2010) pp.243–50, see online at < http://www.pcr.uu.se/research/ucdp/datasets/ucdp_conflict_termination_dataset/>, accessed 8 Mar. 2013.

^16. Online at < http://www.comparativeconstitutionsproject.org/data.htm>, accessed 8 Mar. 2013.

^17. Unpublished databases on constitutional change at the German Institute of Global and Area Studies (GIGA) for the period 2005–10 exist on Latin America and Africa. I want to thank my colleagues who read the relevant country sections of this paper and corrected eventual errors (plus giving more general comments): Matthias Basedau, Sandra Destradi, Sebastian Elischer, Rolf Hofmeier, Claudia Simons, Alexander Stroh, Denis Tull, Andreas Ufen and also Maike Jakusch for some research on background.

^18. The 2001 Constitutional referendum in Guinea was not recorded by the comparative constitutions project. It removed term limits for presidential reelections.

^19. Jennifer Widner, ‘Constitution Writing in Post-War Settings: An Overview’, William and Mary Law Review 49/4 (2008) pp.1513–48, p.1521.

^20. I have excluded from this list the case of Côte d'Ivoire despite its special interest: The temporary suspension of the Constitution's Art. 35 regulating eligibility of presidential candidates in Côte d'Ivoire was a direct outcome of negotiations resulting in the Pretoria agreements in 2005. This enabled the previously excluded candidate Alassane Ouattara to stand in the 2010 elections, a major element of the peace process, but the Constitution was not amended.

^21. The text was very similar to the old Constitution in place before 2009.

^22. Introducing inter alia a two-term limit. This was differently interpreted during weeks of turmoil before the Senegalese presidential election on 26 February 2012 when President Wade stood for a third time as candidate (he was elected a first time shortly before the introduction of the 2001 Constitution; but he failed to win in 2012).

^23. Angola has now a parliamentarian system of government with the president elected by Parliament. In the (undemocratic) Angolan context, this means that the president can be reelected without limitations by a majority that can be taken for granted.

^24. While some elements of a hardline faction continued to stage sporadic attacks.

^25. Don Horowitz, Constitutional Change and Democracy in Indonesia (Cambridge: CUP 2013).

^26. Vandeginste Stef, ‘Power-Sharing, Conflict and Transition in Burundi: Twenty Years of Trial and Error’, Africa Spectrum 44/3 (2009) pp.63–86.

^27. I am not analysing Post-Constitution actor behaviour in this paper. Expectations to usher in an era of both democracy and peace were certainly not met in the Burundian case as first election results have motivated main opposition parties to abandon the multilevel electoral process; a limited upsurge in violence has been recorded since mid-2010. But the Hutu–Tutsi juxtaposition has lost most of its relevance, being replaced by the polarisation between one multi-ethnic dominant party and all other players.

^28. On Rwanda see Alexander Stroh, ‘Electoral Rules of the Authoritarian Game: Undemocratic Effects of Proportional Representation in Rwanda’, Journal of Eastern African Studies 4/1 (2010) pp.1–19.

^29. Denis Tull, ‘Democratic Republic of the Congo’ in Andreas Mehler, Henning Melber and Klaas van Walraven (eds) Africa Yearbook 2, 2005 (Leiden: Brill 2006) pp.217–31. The introduction to the constitution states:

le constituant a structuré administrativement l'Etat congolais en 25 provinces plus la ville de Kinshasa dotées de la personnalité juridique et exerçant des compétences de proximité énumérées dans la présente Constitution. En sus de ces compétences, les provinces en exercent d'autres concurremment avec le pouvoir central et se partagent les recettes nationales avec ce dernier respectivement à raison de 40 et de 60 per cent. (See also Art. 2, 3, 204)

^30. In January 2011 (slightly outside the observation period), the National Assembly revised the Constitution in important ways: Inter alia, presidential elections were reduced to one round of voting, the president was given the ability to dissolve provincial assemblies and to remove governors.

^31. South Sudan authorities have accused the Northern government of interference after the referendum and the situation is not entirely free from violence.

^32. Actually: overrepresentation of Tutsi to guarantee minority rights.

^33. No proportional representation during elections, but composition of Union Parliament and executive with a strong representation principle for the three islands.

^34. According to the DRC constitution, Government on the national and provincial level should be representative (Art. 90, 198). Equitable representation of all provinces is also foreseen for both police and army commanders (Art. 186, 189).

^36. Matthias Basedau, Managing Ethnic Conflict: The Menu of Institutional Engineering (GIGA: Hamburg, GIGA-Working Paper No. 163, 2011).

^35. The necessity to have a decent security sector reform after an incomplete demobilisation attempt in 2007/08, the assassination of a top military leader in 2010 and a further failed coup attempt in 2011 is evident.

^37. Main sources: Africa Yearbook (2004–12), edited by Andreas Mehler, Henning Melber, and Klaas van Walraven (Leiden: Brill); Alex J. Bellamy and Paul Williams, ‘Who is Keeping the Peace? Regionalization and Contemporary Peace Operations’, International Security, 29/4 (2005) pp.157–95, is helpful for peacekeeping operations outside the UN framework (until 2005), UN peacekeeping is well documented in relevant UN webpages. Michael Brzoska ‘Gezielte Sanktionen als Mittel der Konflikteinhegung in Afrika – Erfahrungen und Aussichten, Beiträge aus Sicherheitspolitik und Forschung, 23/4 (2005) pp.209–15, is equally helpful for sanction regimes in Africa until this date; and Matthias Basedau, Clara Portela, and Christian von Soest, ‘Peitsche statt Zuckerbrot. Sind Sanktionen wirkungslos?’, GIGA Focus Global 11 (2010), for sanctions (against authoritarian regimes) in place in 2010 (by USA, UN and EU).

^40. The UNMIN mission to supervise the DDR process consisted of a small unit of retired and unarmed military personnel.

^41. The earlier Indian Peacekeeping mission 1987–90 was a spectacular failure and suffered many casualties.

^38. David Francis and Thomas Kwasi Tieku, The AU and the Search for Peace and Reconciliation in Burundi and Comoros (Geneva: Centre for Humanitarian Dialogue 2011).

^39. Most prominent is probably Michael W. Doyle and Nicholas Sambanis, Making War and Building Peace (Princeton/Oxford: Princeton UP 2007). Failures of UN operations are rather briefly analysed (in comparison to the fully documented successes) in this work. The factors of failure identified in Cyprus and Rwanda are given as the wrong design of the mission and management errors. The authors do not consider analysing local perceptions of mission success.

^42. Simon Massey and Bruce Baker, Comoros: External Involvement in a Small Island State (London: Chatham House 2009).

^43. Whether this speculation can be substantiated remains doubtful at least for the case of Tanzania. President Kikwete may have had a stronger motivation by acting as AU President (personal communication, Rolf Hofmeier).

^44. United Nations Development Programme, Nepal Human Development Report 2009, Kathmandu 2009, p.3.

^45. Mehler (note 8).

^46. See Jacob Bercovitch and Leah Simpson, ‘International Mediation and the Question of Failed Peace Agreements: Improving Conflict Management and Implementation’, Peace & Change 35/1 (2010) pp.68–103 and Kyle Beardsley, ‘Agreement Without Peace? International Mediation and Time Inconsistency Problems’, American Journal of Political Science, 52/4 (2008) pp.723–40.

^47. Hartzell and Hoddie (note 3).

^48. A technically problematic Constitutional referendum was, however, held in Liberia in 2011, slightly outside our observation period.

^49. Donald Horowitz, personal communication. This can be deducted from his idea of ‘asymmetric preferences’, with majorities wanting unimpeded majority rule and minorities attracted by consociational plans, see Donald Horowitz, ‘Conciliatory Institutions and Constitutional Processes in Post-War States’, William & Mary Law Review, 49/4 (2008) pp.1213–48, p.1228.

^50. So political culture could also become a salient context factor for the success of (some elements of) power-sharing, favourable to some and not to other countries. See the argument on the value of socio-political context in René Lemarchand, ‘Consociationalism and Power Sharing in Africa. Rwanda, Burundi and the Democratic Republic of Congo’, African Affairs, 106/422 (2006) pp.1–20, p.3.