ABSTRACT
John Stuart Mill’s scepticism with regard to the possibility of democracy ‘in a country made up of different nationalities’ (Mill 1861: 230) is perhaps the best-known and most widely cited scholarly refl ection of a phenomenon empirically all too often observable as violent ethnic confl ict. Yet, Mill’s scepticism has not, to date, resulted in either ever more homogeneous democratic states or in an inability of heterogeneous countries to become democratic polities. Rather, Mill’s dictum has been taken up as a challenge by scholars and practitioners of institutional design in divided societies to fi nd ways in which democracy and diversity can be married in stable and democratic ways. The answers given in theory and practice are vastly different, and a debate thus continues unabated over which institutional design is best able to provide sustainable democracy in ethnically heterogeneous societies. One such answer is ‘consociational democracy’, prominently associated with the work of Arend Lijphart, as well as more recently with that of John McGarry and Brendan O’Leary. Lijphart began to examine this particular type of democratic system in greater detail for the fi rst time in the late 1960s, when making reference to the political systems of Scandinavian countries and of the Netherlands and Belgium (Lijphart 1968, 1969). He followed up with further studies of political stability in cases of severely socially fragmented societies, eventually leading to his ground-breaking work Democracy in Plural Societies (Lijphart 1977). The phenomenon Lijphart was describing, however, was not new. As a pattern of social structure, characterising a society fragmented by religious, linguistic, ideological, or other cultural segmentation, it had existed and been studied (albeit not as extensively) long before the 1960s. These structural aspects, studied among others by Lorwin (1971), were not the primary concern of Lijphart, who was more interested in why, despite their fragmentation, such societies maintained a stable political process, and identifi ed the behaviour of political elites as the main, but not the only, reason for stability. Furthermore, Lijphart (1977: 25-52) identifi ed four features shared by consociational systems – a grand coalition government (between parties from different segments of society), segmental autonomy (in the cultural sector), proportionality (in the voting system and in public sector employment), and minority veto. These characteristics, more or less prominently, were exhibited by all the classic examples of consociationalism: Lebanon, Cyprus, Switzerland, Austria, the Netherlands, Belgium, Fiji and Malaysia. Some of these consociations – Switzerland, Austria, the Netherlands, and (so far) Belgium – have provided long periods of democratic political stability, while others – Lebanon, Cyprus, Fiji and Malaysia – have not. Lijphart also established conditions
conducive to consociational democracy. These included overarching, i.e. territorial, loyalties, a small number of political parties in each segment, segments of about equal size, and the existence of some cross-cutting cleavages with otherwise segmental isolation. This latter point is important, as the absence of cross-cutting cleavages seems to be a commonality in those countries in which consociation has not succeeded. In addition to Cyprus and Lebanon being examples of failed consociation, Belgium’s most recent experience of protracted government formation brings it close to failure, albeit clearly not with the same violent aftermath that consociational failures in Cyprus and Lebanon had. It is striking to note that in all three examples society had become polarised around a single fault line: respectively an ethno-religious, linguistic and religious cleavage.
