Identity Politics and Elections in Malaysia and Indonesia

Ethnic Engineering in Borneo

Karolina Prasad

• Publisher: Routledge
• Year: 2015
• Print ISBN: 9781138611085
• eBook ISBN: 9781317520276
• Edition: 1
• Topic: Biological Sciences
• Subtopic: Botany
• Index: Biological Sciences

In recent social research, ethnicity has mostly been used as an explanatory variable. It was only after it was agreed that ethnicity, in itself, is subject to change, were the questions of how and why it changes, possible to answer. This multiplicity of ethnic identities requires that we think of each society as one with multiple ethnic dimensions, of which any can become activated in the process of political competition - and sometimes several of them within a short period of time.

Focusing on Malaysia and Indonesia, this book traces the variations of ethnic identity by looking at electoral strategies in two sub-national units. It shows that ethnic identities are subject to change - induced by calculated moves by political entrepreneurs who use identities as tools to maximize their chances of winning elections or expanding support base - and highlights how political institutions play an enormous role in shaping the modes and dynamics of these ethno-political manipulations. The book suggests that in societies where ethnic identities are activated in politics, instead of analysing politics with ethnic distribution as an independent variable, ethnic distribution can be taken as the dependent variable, with political institutions being the explanatory one. It examines the problems of voters' behaviour, and parties' and candidates' strategy in a polity that is, to a significant extent, driven by ethnic relations.

Pushing the boundaries of qualitative research on Southeast Asian politics by placing formal institutions at the centre of its analysis, this book will be of interest to students and scholars of Southeast Asian Politics, Race and Ethnic Studies, and International Relations.

1 Theories, institutions and ethnicity

1.1 Ethnicity and politics

For most of the 20th century, two major themes occupied students of ethno-politics: the first was the theoretical debate on the nature of ethnicity itself and the contents of the concept. The second theme was the practical question of how to design state institutions to eliminate the presumably destructive forces of ethnic diversity. The former problem, typically presented in the literature as a dispute between “primordialists”, who see ethnic identity as single and fixed, and “constructivists”, to whom ethnic identity is multidimensional and fluid, was finally declared as solved by Brubaker, who pronounced primordialism “a long dead horse” (1996, 15).¹ The latter issue of institutional design has been dealt with in a plethora of empirical studies, both qualitative and quantitative, but the results so far have been less than satisfactory, and there is no clear answer to the question of what institutional setting could ultimately minimize the chance of occurrence of conflict in plural societies. The mere assumption that multiethnic states are more prone to conflicts and politically less stable than (relatively) homogeneous states (Horowitz 1985; Rabushka and Shepsle 2009) rests on a premise that is in discordance with the newest findings of empirical studies (Brubaker and Laitin 1998; Davidson 2008a).

Despite the first theme seemingly being closed² and the second hardly lending itself to any hope of solution, this work nevertheless attempts to contribute to the two debates, and does it for two reasons. Foremost, eliminating primordialist elements from the ethno-theoretic discourse did not mean that the constructivist camp agreed on at least the most basic properties and dynamics of change of ethnic identities. Some new theoretical propositions that have arisen from the constructivist triumph over primordialism are promising but they are yet to be tested in confrontation with real-life phenomena. Therefore, the explanatory power of Rogers Brubaker’s (1996, 2004) and Kanchan Chandra’s (2004, 2008, 2012b) concepts, as two of the potentially most powerful conceptual tools in ethnic studies, will be assessed here.

Because, as Brubaker’s argument goes, “ethnic groups” are neither “substantial entities to which interests and agency can be attributed” nor “unitary collective actors with common purposes” (2002, 163), “group” is hardly a useful concept for political analysts. Brubaker’s work has liberated the discussion on identity from the iron cage of “groups” and shifted the analysis to the level of an individual, which paved the way for Chandra’s concept of “ethnic categories”. Therefore, we have arrived at the point where there is no one ethnic identity for each individual, and there is no one ethnic group to which she belongs and acts accordingly – multiple categories of identities exist for each individual, which are invoked in daily life and political behaviour. While this line of thinking is not new (compare Barth 1970), an analytical approach that lends the paradigm to operationalization and comparable studies has recently been systematized and is worth exploring.

The second underlying reason for this study is the consequence of this general agreement that ethnic identity can and does change along with changing incentives. Here the “change” refers to a shift from one category, with which a person identifies (say, “Northerner”), to another category, in which she is also a member (e.g. “Christian”), and at some other time or circumstance to yet another (“Black”), while nominally being a Black Christian Northerner.³ All these categories (and likely a few others) constitute this person’s ethnic repertoire.⁴ So the compelling question is why and how often people shift from one category to another within their repertoire, which in politics translates into a question: how do institutions induce the change and impact its frequency? Consequently, what institutions should states deploy if they face a multiethnic demography? Institutional prescriptions for plural societies developed in the 20th century, while nominally acknowledging the fluidity of ethnic identity, did not take into account the actual possibility of change. Ethnic distribution of analyzed societies was seen as fixed and relied on one arbitrarily chosen (or census-established⁵) dimension, which led authors to provide scientifically developed solutions to the conflict potential in a static, single-dimension polity.

The two states analyzed in this research correspond to the two paradigms of institutional design for divided societies: consociationalism proposed by Arend Lijphart (1969, 2004), and centripetalism⁶ proposed by Donald Horowitz (1985, 1991) and systematized by Benjamin Reilly (2011). One important question is whether these existing recommendations for plural societies succeed in maintaining peaceful and democratic polities, and how they achieve that goal – what are the actual, day-to-day mechanisms of political behaviour that induce non-conflict modes of coexistence between ethnic categories? These designs have a strong standing in political theory and practice and are a ready set of hypotheses and assumptions about patterns of ethnic dynamics in societies in which they function. States frequently deploy propositions embedded in these designs, and comparing ethnic identity change may be one way of assessing how these institutions affect the inter-ethnic relations in these states.

Consociationalism, proposed by Arend Lijphart, in simplest terms is based on power sharing, or “the participation of representatives of all significant communal groups in political decision making, especially at the executive level; [and] group autonomy [which] means that these groups have authority to run their own internal affairs, especially in the areas of education and culture” (Lijphart 2004, 97).⁷ Consociationalism implies that it is institutionally prescribed with which category in their repertoire people should identify, and usually contains all or most of the following: grand coalition governments with all ethnic components deemed relevant involved; proportional representation of different groups in the legislative and in the civil service; segmental autonomy; and a power of veto in case of decisions crucial to ethnic components (Reilly 2006, 815).

Electoral arrangements consociationalists advocate include proportional voting systems and ethnically based parties, and also require “elite-level negotiations between the leaders of the various groups” (Reilly 2006, 815). However, the consociational set-up can involve a variety of different institutional regulations, as Lijphart pointed out in this enumeration of real-life applications:

Broad representation in the executive has been achieved by a constitutional requirement that it be composed of equal numbers of the two major ethnolinguistic groups (Belgium); by granting all parties with a minimum of 5 percent of the legislative seats the right to be represented in the cabinet (South Africa, 1994–99); by the equal representation of the two main parties in the cabinet and an alternation between the two parties in the presidency (Colombia, 1958–64); and by permanently earmarking the presidency for one group and the prime ministership for another (Lebanon).

(2004, 99)

The opposite of consociationalism is centripetalism, which “eschews the reification of ethnic identity inherent in consociationalism and communalism, instead advocating the need for aggregative, centrist and inter-ethnic politics in divided societies” (Reilly 2011, 4). Centripetal-oriented institutions aim to “dilute the ethnic character of competitive politics and promote multiethnic outcomes instead. This means that, for instance, rather than focusing on the fair representation of ethnically-defined political parties, centripetalists place a premium on promoting multiethnic parties and cross-ethnic activity instead” (Reilly 2011, 5). Several institutional incentives were identified for creating multiethnic bridging and political mobilization outside of one’s category. For instance, the goal can be achieved by imposing a ban on ethnic parties, or by the requirement that to win election a presidential candidate needs to obtain his support from not only the majority of voters, but also from voters spread across different ethnic components (usually expressed in geographic terms).

The two cases analyzed in this book correspond to the rationales informing the two approaches political science proposes to the problem of divided societies, although none of them is a perfect realization of the two systems. Malaysia (and Sarawak within it) fulfils several criteria of consociational democracy; power is shared between predefined ethnic categories within a grand coalition and each citizen can claim membership in at least one of the participating ethnic categories; the representation of the ethnic components in the legislative is divided more or less proportionally, although the proportionality is not a result of proportional elections (generally advocated by consociationalism), but results from assigning constituencies/seats to each ethnic category. Some autonomy in cultural affairs is granted to ethnic categories,⁸ and the element of elite negotiations is very strong.⁹Consociational design should be expected to reify and arrest ethnic identity for extended periods of time: power is shared and negotiated between the same categories over and over again (e.g. between Muslims and Christians, Flemish and French speakers, Blacks and Whites), and there is little space for activation of other categories. Sarawak in Malaysia, therefore, is used here as a case study to test whether and how ethnic categories are indeed arrested in a consociational polity, or if the ethnic identity change still happens in this society, thanks to the institutions in which it takes place.

Indonesia and one of its provinces selected for this study, West Kalimantan, broadly correspond to the centripetalist design. Along with direct presidential elections in 2004, the country introduced a stipulation for winning presidential candidates not only to obtain the majority of votes, but also to gain support from at least 20% of voters in at least half of all the provinces in Indonesia, along with the ban of regional parties (which as a result also bans regional ethnic parties), a proportional electoral system and incentives for parties to build coalitions to nominate executive candidates. Ethnic mobilization is frowned upon with one important exception – religion, specifically Islam. Parties calling for more Islamic influence in the state operate freely and individual candidates (also from non-Islamic parties) involve religious elements in their campaigns. This way, Indonesia fulfils most – but again, not all – of the criteria to be recognized as a centripetalist design. Although centripetalist institutions are designed to divert political discourse away from ethnically driven interests, this has not been so far achieved in Indonesia; politics is very much focussed around ethnicity, all the more so at the sub-national level. The search for a mechanism of politicization of ethnicity within a seemingly centripetal setting is one of the tasks of this research.

In an effort to overcome the centripetalism versus consociationalism debate, Kanchan Chandra followed through with her theoretical propositions and, assuming multiple and fluid ethnic identities for each individual, she put forward a pragmatic argument: fluidity of ethnic identities offers a potential for an alternative institutional design for plural societies. Chandra suggests that the key is to induce such a combination of institutions that leads an individual to frequently shift between her identity categories. Simply put, institutions should be such that they neither prescribe identifying with a particular category (as it is in consociationalism), nor should they get in the way of ethnic mobilization (2008, 27). Chandra named the process an “ethnic invention”, as it encourages political activation of ethnic categories that otherwise might remain only nominal and “create[s] incentives for voters to retain multiple identities in their repertoires” (2008, 27). Moreover, “ethnic invention” is conditional on lack of constraints for ethnic mobilization in a polity (e.g. ethnic party bans).

Simultaneously, therefore, Malaysia and Indonesia will be test cases for the question hidden behind Chandra’s (2008) proposition of an alternative institutional approach for divided societies. Chandra’s proposition implies that each election can be and should be an opportunity for a voter to identify with a different identity from her ethnic repertoire, and hence, the more elections in which the voter participates, the more activated (i.e. politically relevant) identities she retains in her repertoire, and the more stable the polity is. Malaysia and Indonesia are excellent case studies to test this hypothesis: a state with few elected offices (Malaysia holds a total of only two elections: to the state and federal legislatures), and a state with a total of seven elections for different offices (Indonesia has direct executive and legislative elections on three tiers) will be compared to see in which of them voters retain more categories in their ethnic repertoires. As we saw earlier, however, Malaysia places no obstacles on ethnic mobilization, which makes ethnic mobilization easier, while Indonesia limits opportunities for ethnic mobilization through its centripetal institutions. We will ask: in which of the two studied cases is a more frequent and faster shift between identity categories observed? And to which particular elements of the institutional maze should the identity shift (and lack thereof) be attributed?

Being committed to analysis of ethnic identity change, we do wish to ask: what are the ways of inducing or hampering ethnic identity shifts under these three institutional designs? The three designs have specific goals, which, if achieved, would lead to different outcomes. Consociationalism wants to achieve democratic stability through the long-term arrest of ethnic categories (under the implications of this system each individual is induced to identify with only one category over long periods of time and in all political circumstances). Centripetalism, on the other hand, attempts to make all ethnic categories politically irrelevant as it discourages mobilizing of any ethnic categories. Chandra’s “ethnic inventionism” seeks to disperse the ethnic loyalty of each individual between several categories. By addressing these issues, we will arrive at some interesting conclusions regarding the stability of political and social life under the different settings. West Kalimantan has experienced several violent episodes, including recent and deadly ones, and this experience is assumed to factor in day-to-day political proceedings. Sarawak has remained peaceful over the decades despite the 1969 turmoil in Peninsular Malaysia, and we need to ask what contributed to Sarawak’s social stability.

Let it be underscored that the distinction here between consociationalism, centripetalism and “ethnic inventionism” serves the purpose of establishing the research design, as it offers a coherent theoretical approach to the question of institutions and ethnicity and helps generate hypotheses. Although these institutional designs offer a ready-made set of 0, 1 value variables and differentiate neatly between the cases, the actual analysis of this work will look into each element of the institutional system as a distinct variable. A close observation of two cases over a period of time will allow capturing the impact of each of the institutional elements on ethnic identity change. Hypotheses...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of figures
• List of tables
• 1 Theories, institutions and ethnicity
• 2 Sarawak: institutional and historical overview
• 3 Ethnic identity change in a consociational polity
• 4 West Kalimantan: institutional and historical overview
• 5 Identity activation in the new institutional setting
• 6 Sarawak and West Kalimantan: a comparison
• Appendices
• Index