An exploratory study of constitutional design in three island states: Seychelles, Comoros, and Mauritius

ABSTRACT

This article explores the effects of constitutional design on conflict management and democratisation outcomes in three African island states: Seychelles, Comoros, and Mauritius. As an exploratory framework, the article reflects upon Arend Lijphart’s theory of accommodative constitutional design in view of political developments in the three cases, beginning around their independence and continuing into recent times. The case studies provide some general support to Lijphart’s accommodative theory, and particularly to the need to facilitate broad representation and an overall sense of political inclusion, but they also raise critical questions regarding the effects of his institutional recommendations. Contrary to Lijphart’s approach, majoritarian legislative elections might under certain conditions facilitate greater inclusion than proportional representation, and direct presidential elections could sometimes be more inclusive than a parliamentary executive. The findings on Lijphart’s recommendations of group autonomy are less clear, particularly regarding the question of whether territorial autonomy might increase risks of secession.

KEYWORDS:

• Constitutional design
• Lijphart
• Comoros
• Mauritius
• Seychelles

Acknowledgments

Many thanks to Professor Alan Kuperman for invaluable discussion and comments on several revisions of this paper. Thanks also to the two anonymous reviewers, and to the editors of this journal, for their generous and extremely helpful feedback. Remaining faults are of course the author’s alone.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Eliezer S. Poupko received his Ph.D. in Public Policy in May 2017 from the Lyndon B. Johnson School of Public Affairs, University of Texas at Austin. His research interests include constitutional design, election law and policy, and democratic theory. His previous publications are: ‘Africa’s Domestic Institutions of Integration and Accommodation: A New Database’, in Constitutions and Conflict Management in Africa: Preventing Civil War through Institutional Design, edited by Alan J. Kuperman, 183–224 (University of Pennsylvania Press, 2015); and ‘Reinterpreting p: A New Theory of How Individual Votes Contribute to Electoral Outcomes’. Election Law Journal 14 (2): 111–135 (2015).

Notes

1 But see Baldacchino (2012) and Veenendaal (2013), finding that small island states may also be subject to personalistic politics and autocratic tendencies. For a survey of the literature on small island states and democracy, see Srebrnik (2004), who also discusses possibilities of instability and identity-based conflict. The notion of a negative relationship between size and democracy is a long-standing one, tracing back to ancient Greek thought (Dahl and Tufte 1973, 4–5). Lijphart also discusses how smallness may be favourable to democratic development (Lijphart 1977, 65–70). One might question whether Mauritius – with a population over one million – should qualify as a ‘small’ island state (Anckar 2010), but it is generally included in lists of these states (e.g. UNESCO 2016).

2 According to Lijphart, consensus democracy “tends to emphasize formal-institutional devices,” while consociational democracy “relies to a larger extent on informal practices” (Lijphart 1989, 40).

3 Lijphart does not explicitly mention consensus democracy in his 2004 article, and he mentions the consociational model only in passing, but his recommendations are generally consistent with both models. The full model of consensus democracy includes some elements not addressed in the recommendations for divided societies, including a ‘corporatist’ (versus pluralist) interest group system, bicameralism, rigid constitutionalism with strong judicial review, and an independent central bank (Lijphart 2012, 3–4). The full consociational model includes the institutional elements of a grand coalition and a minority veto (Lijphart 1989, 40).

4 Horowitz indicates that his approach should also be classified as a form of power sharing and accommodation (Horowitz 2002, 23; 2008, 1216). See also Sisk (1996, x), indicating that integrative approaches like Horowitz's may also provide power sharing. For an empirical study of centripetal design, see Reilly (2001).

5 To be fair, Lijphart states, ‘Within the category of majoritarian systems, a good case could be made for Horowitz's alternative-vote proposal, which I agree is superior to both the plurality method and the two-ballot majority runoff.’ However, he goes on to assert the existence of a ‘scholarly consensus against majoritarian systems in divided societies’ and to question any deviation from PR (Lijphart 2004, 100). In other work, Lijphart states that Horowitz's vote-pooling recommendations ‘do not differ a great deal from majority-rule democracy’, and he criticizes Horowitz's view of incentives for moderation under the alternative vote (Lijphart 2002, 47–48).

6 The title president does not necessarily imply presidentialism, which is defined by direct popular election of the head of government.

7 Freedom House (2017) classifies Seychelles as ‘free’ in 1976, ‘partly free’ in 1977–1978, then ‘not free’ from 1979 to 1991.

8 In April 2016, the constitution was amended to limit the president to two five-year terms.

9 From 1992 through 2016, Freedom House (2017) classifies Seychelles as ‘partly free’.

10 2001 Constitution of Comoros, Article 13.

11 2001 Constitution of Comoros, Article 20.

12 Freedom House (2017) classifies Comoros as ‘partly free’ since 1991, before which it was classified as ‘not free.’

13 The system of plurality voting in multi-member districts still in use in Mauritius, in which voters cast one vote for each available seat, is often called block voting.

14 1968 Constitution of Mauritius (amended through 2010), Chapter VI-A, Articles 75A-75E.

15 It is perhaps concerning, however, that all prime ministers of Mauritius have been members of the Ramgoolam or Jugnauth families, other than the few years in which Paul Bérenger held the position.

16 Selway also indicates that PR-based modifications to the electoral system could risk removing incentives for inter-communal coordination (Selway 2015, 175–178).

17 Brautigam refers to the best loser system as a ‘consociational compromise’ (Brautigam 1999, 147); see also Srebrnik (2002, 289).