Consociationalism in the Post-Colonial World

Consociationalism in the post-colonial world

A comparative study of Fiji and Mauritius

Gabriel Lönn

Bachelor Thesis, Autumn 2020 Department of Government Political Science C Supervisor: Johanna Pettersson Words (Pages): 13980 (36)

Table of Contents Introduction 3

Aim and research question 4

Prior research and theory 5 Democracy in post-colonial states 5 Mauritius and Fiji 5 Consociational Democracy 7 Consociationalism in a post-colonial context 7 Research design and Method 7 Design 7 Method 11 Response to criticism 13 Material 13

Analysis 14 Mauritius constitution 14 The first period (1968 - 1992) 14 The second period (1992-2020) 15 Fiji constitution 16 The first period (1970 - 1992) 16 The second period (1992-2020) 17 Mauritius in practice 19 The first period (1968-1992) 19 The second period (1992-2020) 22 Fiji in practice 24 The first period (1970-1992) 24 The second period (1992-2020) 26 Results and discussion 29 Analysis summary and interpretation 29 Other possible explaining variables 30 Future research 30 Conclusion 31

References 32

Appendix 35 Mauritius Prime Ministers 35 Fiji Prime Ministers 36

Introduction The 1960s and 70s saw the birth of many new nations as a consequence of the former European empires - Britain, France and Portugal - letting go of their former colonial holdings in Africa, the Caribbean and the South Pacific, sometimes voluntarily but other times due to violent independence struggles. This significant period in history raised many questions in regard to how these newly formed nations would fare both politically and economically; in particular since many of them consisted of peoples that had never been part of the same society before colonial rule and were now forced to coexist within the same borders. Now some 50 years later it can be concluded that this process has been at least partially successful, with some nations managing to develop functioning structures to handle internal conflict through democratic means rather than violent ones - some scholars even regard this period as a third wave of democratization (Huntington, 1991). Other states have not been as successful, however, and have instead entered into a spiral of authoritarian rule, civil conflict and the suppression of democratic rights. This study will seek to further explore this discrepancy of why some countries manage this democratic transition while others do not.

1 Mauritius is an example of a post-colonial nation that has managed to develop a functioning democratic system in a relatively short period after its independence. This small island nation to the east of Madagascar gained its independence from the United Kingdom in 1968 under unusually calm and peaceful circumstances. After an initial setback in which the election of 1972 was cancelled due to a “state of emergency” declared by the government, the country has successfully held eleven national elections between 1976 and 2019 - with three different parties having held the post of Prime Minister at least once. Today, Mauritius is frequently ranked as the most democratic nation in Africa and among the most democratic developing 2 nations in the world . What makes it stand out even more, however, is the fact that the nation and its society since the inception has been very heterogeneous, with a wide variety of different ethnic backgrounds, religions and languages all being represented in the population. This is a consequence of Mauritius colonial heritage, since the majority of the population are descendants of people brought to the island either as slaves or later as indentured workers in order to work the sugar fields. While a diverse society with no majority group and many large subdivisions is usually seen as an impediment to democracy, this has not been the case in Mauritius, which makes it an interesting object of study.

Fiji, on the other hand, is an example of a post-colonial nation that has developed in a considerably less democratic direction. The initial prospects looked positive; after gaining independence from the United Kingdom in 1970, the nation had a 17-year period of prosperity, when it thrived both economically and democratically. This period came to an abrupt end in 1987, when the nation saw its first change of power due to a democratic election. Less than a month into the mandate, a military coup took place which deposed the government; this was followed by a second coup four months later in which the British

1 Note that the term “post-colonial” in this paper will be used synonymously with “previously colonised” and thus has no direct links to post-colonial theory 2 See for example The Economist’s Democracy Index, where Mauritius ranks 18th among the 167 nations ranked, placing it in the category of “Full Democracy” (EIU, 2020).

monarchy was abolished and a republic was established. Two more coups would follow in the next 20 years, a civilian coup in 2000 and yet another military coup in 2006. Lately, the country has been advancing in a more democratic direction but international observers still testify about how the government is committing acts of voter suppression, attacks on 3 independent media and has suspended or even imprisoned its political opponents . Much of the political turmoil that Fiji has experienced can be seen as a consequence of ethnic and cultural divides in the population; similar to Mauritius, a large share of the Fijian population descend from individuals that were brought to the islands as indentured workers during the British colonial rule. Unlike Mauritius, however, Fiji has not managed to handle these issues within the boundaries of democracy, which in turn has led to the repeated pattern of military coups and subsequent restructurings of the political system. This makes Fiji an interesting object of study in its own right, but also a suitable object of comparison in relation to Mauritius.

Aim and research question This paper seeks to understand why some post-colonial nations with plural societies manage to contain these cleavages within the realm of democratic politics while other nations fail to do so and instead develop oppressive and sometimes violent systems where one or several groups benefit at the cost of others. This is on one hand relevant within the field of democratic research, since traditionally deep cultural divides within a society is seen as purely being detrimental to the development of democracy but as Mauritius shows, this is not always the case. At the same time, the question has a practical relevance since an expanded knowledge on how some nations manage this balancing act may be beneficial to other nations dealing with similar issues.

For the purpose of researching this broader topic, this paper will focus on Mauritius and Fiji as cases of successful and failed democratization respectively. The research question chosen for this purpose is therefore:

- Why has Mauritius developed a functioning democracy when Fiji has failed to do so?

The theory that will be used in order to find the answer to this question is Arend Lijphart’s theory of consociational democracy, which describes a certain type of political system that only exists in some plural societies, that puts emphasis on cooperation between different groups and broad consensus in political decision-making. The hypothesis is that Mauritius political system would more closely resemble Lijphart’s model for a consociational democracy than Fiji’s and that this in turn would explain the difference in democracy between the countries. The theory itself will be further explained in a later section, as will the method for verifying the hypothesis.

3 For a comparison with Mauritius, see The Economist’s Democracy Index where Fiji ranks 81st among the 167 nations ranked, placing it in the category of “Hybrid regime” (EIU, 2020).

Prior research and theory

Democracy in post-colonial states One of the most famous traditional theories on why some nations develop democracy while others do not is Modernization theory, which in simplified terms state that sustained economic growth in a society will propel the evolution of democracy forward; this connection was perhaps most famously made by Seymour Martin Lipset in his 1960 book Political Man. Since its inception, the theory has seen much support, but also much criticism; it should nevertheless not be ignored considering the massive impact that it has had on the field.

Since the latest period of decolonisation in the 1960s and 70s, much scholarly attention within the field of democratic research has been dedicated to the study of post-colonial societies, particularly in Africa. Chimakonam, Agu and Agbo (2014, pp 137-155) argue that the export of Western ideology and capitalism to Africa has been detrimental to the development of democracy on the continent, since these values are difficult to reconcile with the way that African societies traditionally has been governed, through deliberation and unanimity in decision-making. While the authors themselves do not bring up the topic, this is still interesting to relate to the concept of consociationalism since one of its central tenets is to build consensus in decision-making, in order to make decisions with broad support among many different groups. This system of governance may thus be more suited to certain post-colonial contexts - such as that of Sub-Saharan Africa - than more adversarial systems modeled on British or American politics.

Witsoe (2011, pp 619-631) instead focuses on the case of democracy in India and argues that democracy in modern society can not be analyzed with a single framework that fits all cases; this is especially true in post-colonial societies. Witsoe illustrates this by using his experiences from observing a local election in the Indian state of Bihar, which he dubs as an “alternate democracy”, that needs to be understood in the historical context which has shaped it, from post-colonial times, through colonization and up until present day. This focus on historical processes should also be kept in mind when performing the analysis in this paper.

Mauritius and Fiji The unique case of Mauritius has piqued the interest of many political scientists over the course of its 50-year lifespan. Using a wide array of approaches they have all attempted to explain the so-called “Mauritian Miracle”; how this small island nation with its divided society has achieved such economic and democratic success in a relatively short time. Carroll and Carroll, in one article from 1999 (pp 179-197) directs their attention on how, among other factors, the integration of cultural and religious minorities in the political sphere and the accommodation of the interests of these communities has been crucial to democratic consolidation. In another article by the same authors from the ensuing year (2000, pp 120-142), they discuss further how the Mauritian government has been open to listening to the opinions of the plural civic society before making political decisions and, of note for this

paper, how Mauritius has features of the consociational democracy described by Lijphart and other consociational theorists. They do not, however, analyze more in-depth what this has meant for Mauritius democratic development. Lastly, Bräutigam (1997, pp 45-62) focuses her attention to the institutions and policies that have laid the foundation for democratization in the nation; interestingly, she refers to Mauritius political system as a case of “modified consociationalism”, though she does not explain precisely in which ways it is consociational. This along with the references to consociationalism in the article by Carroll and Carroll makes it even more relevant to analyze more in-depth whether Mauritius fits the bill of consociational democracy.

Though not for the same reasons, there has been a similar scholarly interest in Fiji; particularly since the democratic system failed in 1987 after two consequent military coups. McCarthy (2011, pp 563-578) attempts to explain the failure of democracy in Fiji by looking at the political role of the Fijian military, traditional institutions developed during colonial rule - such as the Great Council of Chiefs - as well as grievances among iTaukei, Fiji’s indigenous population. The author finds that Fiji's society is deeply divided between the iTaukei whose ancestors have lived on the islands since long before colonial rule and the share of the population who came to the islands during this era - primarily from the Indian subcontinent but also from Europe, China and other Polynesian islands. This finding is shared by Lawson (2012, pp 293-315) who describes Fiji as an “ethnic democracy” where voting historically has been determined by ethnicity rather than politics; for example, during the first 17 years of independence, indigenous Fijians primarily voted for the governing Fiji Alliance Party(FAP) while Indo-Fijians tended to vote for the Fiji Labour Party (FLP). This “ethnification” of politics was an important factor as to why the victory of the FLP in the 1987 elections was seen as threatening by many indigenous Fijians. It should be noted here that parties representing different ethnic, cultural or religious groups are a common feature in consociational democracies, which makes it further intriguing to see if Fiji has other elements of this system. In an article from 2007, Iyer (pp 127-153) states that the 1997 Fijian constitution was an attempt to break the hegemony of the iTaukei and bring about more political equality between different ethnic groups. The author describes this as an attempt at “enforced consociationalism”, which he believes may have had an adverse effect on the development of democracy by being one of the driving factors behind the 2000 civilian coup; this coup took place after the FLP once again came to power after the 1999 elections. The conclusion by Iyer is that consociationalism is better suited for societies where the cleavages between groups are limited unlike in Fiji’s case where they are deeper and significantly harder to bridge, a finding that certainly deserves further examination.

Finally, Carroll (1994, pp 301-324) makes a comparison between Mauritius and Fiji and more specifically how Mauritius has managed to contain ethnic conflicts from erupting into violence, while Fiji’s conflicts has led to military intervention and the dissolution of democracy. He finds that the main difference between Mauritius and Fiji is that the latter has an indigenous population that claims to have a special protected status and thus their interests are put ahead of those of other groups in society. Mauritius, meanwhile, has no “native” population since the island was uninhabited until the Dutch settled it in the late 16th century.

While Carroll’s question resembles the one in this paper, his theoretical framework is considerably different as it focuses on perceived entitlement by one group in society. Additionally, his article is 26 years old and both nations have seen considerable changes since then; in Fiji’s case two additional coups and two re-writings of the constitution. Carroll’s findings will nevertheless be kept in mind for the analysis, and will be discussed further in the final section of the paper.

Consociational Democracy The main theoretical framework that will be used in this paper is derived from Arend Lijphart’s works on consociational democracy, and more specifically his 1977 book Democracy in Plural Societies: A Comparative Explanation. In the second chapter of the book, Lijphart outlines the four features he regards as being characteristic of a consociational democracy; these will also form the basis of the analysis in this paper (Lijphart, 1977, pp. 25-52). The first characteristic is according to Lijphart the most essential; the nation should be governed by a “grand coalition” of members representing all the significant social or cultural subdivisions within the country. This is perhaps most intuitive to think of as a broad coalition government, where the cabinet consists of members from several different political parties, each representing the interests of one of the principal communities. The remaining three features are each given equal importance by Lijphart: a mutual veto for all the minorities in the grand coalition, proportionality in the distribution of both political posts and public means and lastly autonomy for each individual group to handle their own internal affairs. All these four elements will be further expanded on and operationalized later in the paper.

Consociationalism in a post-colonial context Lijphart’s theory of consociationalism stems from his research into European parliamentary systems in the late 1800s and early 1900s, more specifically four countries; Austria, Switzerland, Belgium and his own native country, The Netherlands. According to Lijphart himself, however, the theory is applicable even outside of a European context, and a later chapter in the book is fully dedicated to discussing consociational democracy in a third-world scenario (Ibid, pp 142-176). In this chapter Lijphart discusses instances where former colonial nations have attempted to implement consociational systems, with varying degrees of success. Malaysia and Lebanon are presented as two cases where this mode of governance has been successfully used for an extended period of time, while Cyprus is presented as a case where this was attempted but the implementation ultimately was unsuccessful. Lijphart explains this by the fact that the conditions in Malaysia and Lebanon were more favourable for the success of consociationalism than the ones in Cyprus.

Research design and Method

Design As has been alluded to already, the analysis in this paper will take the form of a comparative case study of Mauritius and Fiji as cases of post-colonial nations with plural societies. These

two nations have been selected for the purpose of using a Most Similar Systems Design, where they are meant to resemble each other as closely as possible on all relevant variables except the dependent and independent variables that will be analyzed. The dependent variable in this case is the level of democracy in the state, while the independent variable is the degree to which each state lives up to an ideal type of consociational democracy, based on Lijphart’s definition. For the sake of avoiding misconceptions, it should be clarified here that these two variables are completely separate concepts from each other; level of democracy refers strictly to a measurement of how democratic a state is while consociational democracy is one type of democratic system used by some states. Further, consociational democracy is in no way a prerequisite for democracy considering the majority of democracies are not consociational in nature; consociationalism is simply one way in which plural societies can manage internal quarrel in a non-violent manner.

The democracy variable in this paper will be retrieved from The Economist’s yearly Democracy Index, which ranks countries using a scale from 0 to 10 and afterwards sorts them into four categories - Full Democracy, Flawed Democracy, Hybrid Regime and Authoritarian (EIU, 2020). This ranking is based on five factors that together aim to capture what the authors behind the index consider to be the most crucial aspects of a liberal democracy; electoral processes and pluralism; civil liberties; the functioning of government; political participation and lastly political culture (Ibid., pp 51). In total, the index ranks 165 out of the 194 nations recognized by the UN, plus Taiwan and Palestine. Out of these 167 nations, Mauritius ranks 18th with an index of 8,22. placing it in the category of “Full Democracy” and notably ahead of old, established democracies such as France (20th) and the United States (25th). It is also the highest ranked country in Africa and the second highest ranked developing country, after Uruguay (15th). Fiji, on the other hand, receives an index of 5,85 and a ranking of 81st, placing it in the category of Hybrid Regime. It should be noted, however, that this is still a significant increase from 2013, when the nation received its lowest 4 marking of 3,61 which would have placed it in the category of “Authoritarian” . While the primary indicator of “level of democracy” in this paper is the EIU Democracy Index, this can be complemented by looking at another well-renowned source - Freedom House - which ranks Global Freedom in 210 countries and territories based on two factors; political rights and civil liberties. In this index, which spans between 0 and 100, Mauritius receives a score of 89 while Fiji gets 60, further indicating that there exists a significant difference between the two countries.

The explanatory variable, adherence to the ideal type of consociational democracy, will be discussed extensively in the next section of the paper along with the method that will be used to determine this variable. The remainder of this section will be dedicated to further exploring the similarities between the two countries, which have only been hinted at thus far. Firstly, the nations are of similar size in terms of population; Fiji has a population of just under 900 000 while Mauritius has about 1,27 million (WPR, 2020). The difference is larger when it

4 The Economist’s Democracy Index only dates back to 2006, meaning that there is no way to see how Fiji ranked during its 17-year grace period before the first coup, nor during the turbulent period between 1987 and 2006.

2 comes to the actual size of the nation; Fiji has an area of about 18 000 km while Mauritius 2 area is about 2 000 km . In an international context they are both small nations however, and there is no theoretical ground for believing that this difference in size would affect the level of democracy. Further, both nations have societies that can best be described as plural, both when it comes to ethnicity and religion. In Mauritius between 65-70% of the population is of Indian descent, followed by about 25-30% descending from Africa and about 5-10% 5 belonging to another ethnicity, mainly Chinese or European . Religiously, 49% are Hindus, 27% are Catholics, 17% Muslims and 6% belong to another branch of Christianity (CIA, 2020b; WPR, 2020b). Fiji, meanwhile, has a population where 57% belong to the native population, iTaukei, 38% are of Indian descent, and 5-6% belong to a different ethnic group, mainly European, Chinese or Polynesian. When it comes to religion, 45% of the people are Protestants - with the Methodist church being the largest religious congregation - 28% are hindus, 9% are Catholics, 6% are Muslims and approximately 10% belong to another Christian branch than the two mentioned (CIA, 2020a; WPR, 2020a).

As has been mentioned already, both nations gained their independence from the United Kingdom around the same time, and both has since then become republics - Fiji in 1987 and 6 Mauritius in 1992 - while still remaining part of the British Commonwealth . Both are also unitary parliamentary systems with a single chamber; in Mauritius case the main island is divided into 20 constituencies that each elect three members for parliament while the semi-autonomous island of Rodrigues elects two members. In Fiji, on the other hand, the whole nation since 2013 serves as a single constituency with proportional representation based on the percentage of votes received by each party. This difference between the two countries will be discussed in further detail in the analysis, but it is not enough by itself to completely dismiss the comparison.

In regard to their level of development, the nations are also quite similar; Mauritius Human Development Index was 0,796 in 2019 while Fiji’s was 0,724, placing both nations in the category of “High Human Development”, the second highest of four categories (UNDP, 2019, pp 314-317). The same can not be said of their economic development since Mauritius in 2019 had a GDP/capita of $11 200 while Fiji’s was only $6 200 (World Bank, 2020). One may argue here from a modernization perspective that Mauritius democratic success has been caused by its higher degree of economic development, while the opposite is true in Fiji. In order to examine this argument properly, it would thus be relevant to look at the GDP/capita 7 of the two nations shortly after their independence. In 1976 Mauritius had a GDP/capita of $780 while Fiji’s was $1 180, which according to modernization theory would indicate that at the outset Fiji actually had better prospects for developing democracy; it would thus be more reasonable to assume that the difference in economic development today is caused by Mauritius being more democratic and not vice verse. Interestingly, the first year in which

5 These numbers are based on approximations, since official censuses by the Mauritian government no longer collect information on ethnicity. 6 Fiji did leave the Commonwealth in 1987 but later rejoined. It has since been suspended twice, following the coups of 2000 and 2006 but is today once again part of the Commonwealth. 7 This year was chosen since it is the first year for which the World Bank has statistics of Mauritius GDP/capita.

Mauritius GDP/capita surpassed Fiji’s was 1987 (see Figure 1) - when Fiji’s democracy saw its first real setback - and it has since then continued to climb at a faster rate than that of Fiji.

Figure 1

Data from the World Bank (2020)

Table 1 summarizes all the factors that have been discussed thus far, and shows that the two nations are in fact very similar except for the two factors that are of interest for this analysis; the Most Similar Systems Design would therefore be appropriate to utilize.

Table 1 Fiji Mauritius

Independence 1970, from United Kingdom 1968, from United Kingdom

Republican Status 1987 1992

Member of the Yes Yes Commonwealth of Nations

Plural society Yes Yes

Population at independence 521 000 799 000

Population (2020) 899 000 1 273 000

Size 18 274 km2 2 040 km2

Human Development Index 0,724 0,796 (2018)

GDP/Capita (1976) $1 182 $777

GDP/Capita (2019) $6 220 $11 204

Electoral system Unitary, parliamentary, with Unitary, parliamentary, with a single constituency 21 separate constituencies

Democracy Index (2019) 5,85 (Hybrid regime) 8,22 (Full Democracy)

Freedom House (2019) 60 (Partly Free) 89 (Free)

Elements of consociational Fewer More democracy (Hypothesis)

Method The method that will be used in this paper is a form of analysis based on ideal types, in order to determine how much each object of study resembles Lijphart’s definition of a consociational democracy. As has been mentioned above, this definition consists of four different elements; these will also form the basis of the analysis in this paper. These four elements, as a brief repetition, are as follows: 8 (1) A grand governing coalition consisting of representatives of all the sections of society, (2) A mutual veto for all minority groups that take part in the coalition, through which these actors can protect the interests of their respective communities, (3) Proportional representation for all societal groups on both political posts and civil service positions. Additionally, proportionality should also be applied to the distribution of public means, (4) Autonomy for all sections of society to handle their own internal matters without outside interference. While his definition is theoretical in nature, Lijphart himself uses it in an operational manner later in the book, when discussing examples of consociationalism in the third world (Lijphart, 1977, pp 142-176). This would be one argument for not making significant alterations when operationalizing the theory and thus this analysis will use a number of questions based on these criteria in order to determine how much each nation resembles the ideal type.

The analysis itself will be divided into four separate sections based on two dimensions. The first of these dimensions is one of constitutional conditions versus actual practice. This distinction is important since a constitution may include consociational elements that are not utilized in practice in the same way that consociationalism may develop organically without any constitutional basis. To get the full picture it is thus necessary to observe both these aspects. The questions used for the analysis will thus be as follows:

8 It should be clarified here that societal “sections” or “groups” in this context refers strictly to ethnic/cultural/religious/lingual groups and not groups related to, for example, age or gender.

In regard to the constitutions

Grand coalition 1. Does the constitution as it is constructed allow for a plurality of parties in the parliament?

Minority veto 2. Is there a constitutional mechanism through which minority parties in Parliament can veto decisions that would potentially be harmful to their interests?

Proportionality 3. Are there constitutional structures that guarantee proportional representation for the different sections of society on political and civil service positions?

Autonomy 4. Does the constitution allow for all societal groups to autonomously handle their own internal affairs?

In regard to actual political practice

Grand coalition 5. Is there a political history of parties working together in broad coalition-governments?

6. Is there a non-governmental, counseling body that consists of all major sections of society and that advises the government on important political decisions?

Minority veto 7. In coalition governments, are there occurrences where minority parties have vetoed or threatened to veto decisions that majority parties have attempted to implement?

Proportionality 8. Is the composition of the parliament and other governmental bodies representative of the population as a whole or are any groups severely over- or underrepresented?

Autonomy 9. Do the different segments of society have the ability to manage their own internal issues outside the influence of other segments or the central government?

The second dimension that will be used to divide the different parts of the analysis is time; Fiji especially but also Mauritius to a lesser degree, has seen significant changes in their political systems over the course of their 50-year independence and it is thus necessary to observe them over time in order to get the full picture. Further, the time-dimension also adds

a possibility to determine causality, by observing if consociational elements in the first period have benefitted democracy down the line. For this purpose, the time-span has been divided into two periods, the first being from independence until 1992 and the second being from 1992 until the present year, 2020. The year 1992 was chosen since it is roughly in the middle of the total time-span and since this year saw significant political changes for both nations; in Mauritius case a transition from monarchy to republic and in Fiji, the first democratic election after the 1987 coups and the new constitution of 1990.

Response to criticism One potential weakness in the method is its intersubjectivity since some of the questions described above out of necessity are quite broad and may leave some room for interpretation by the researcher. In order to avoid this as far as possible, the historical analysis of the nations will focus primarily on major events during the two periods while simultaneously seeking to not leave out any events that may have significantly changed the answer to the question. By taking these measures, the study should be replicable by other researchers without reaching widely different conclusions.

Material For the section relating to the composition of the constitution, the principal material will consist of the constitutions of the two states. In Mauritius case this is a single document since Mauritius has not changed its constitution since independence; it has however been amended on several occasions since then and thus the book Parliament in Mauritius (Mathur, 2017) will be used to track how it has evolved over time. In the case of Fiji the constitution has been replaced three times since the original independence document from 1970 - in 1990, 1997 and most recently in 2013. Analyzing all four of these documents in-depth would not only be a big undertaking time-wise, it would also mean that Fiji was given an unproportional amount of attention in the analysis; the decision has thus been made to limit the scope to mainly discuss the oldest and newest constitutions, with the 1990 and 1997 constitutions instead being discussed in regard to the actual practice.

For the section on political practice, a wider range of material will be needed. Some of this material will be biographical in nature, chronicling the political history of the two nations from pre-independence until more recent political developments. Previous research into the political systems of the two nations, such as that described in a previous chapter, may also serve to illuminate what political practices are common. Additionally, this material will be complemented by using reports from NGOs and journalistic sources from both local and international media.

Analysis

Mauritius constitution

The first period (1968 - 1992) (1) The 1968 constitution of Mauritius ascertains that the nation will consist of 20 constituiences which each will send three members to the National Assembly and one 9 constituency, the island of Rodrigues, which sends two members . Each voter in a given constituency votes for exactly three candidates and the three candidates with the most votes get the three seats for that constituency (Mauritius, 1968, Schedule 1); this can be described as a modified version of the British “first-past-the-post” system, which normally operates in single-member constituencies. While this system is far from being fully proportional, it does open up the possibility for independent candidates or candidates from smaller parties to make their way into the Assembly, provided they are at least the third most successful in their given constituency. Another mechanism that may secure a wider range of parties is a system that has later become known as the “Best Loser”-system, which distributes eight additional seats to candidates who were close to gaining the required number of votes and who belong to ethnic/religious communities that were underrepresented among the first 62 members elected (Ibid.).

(2) The constitution makes no mention of veto power for any members of the Assembly or certain parties. The only person with any form of veto-power is the Governor-General, who under specified conditions is able to withhold his or her assent for bills passed by the Assembly; officially on behalf of the monarch of the United Kingdom (Ibid., Section 46). The only form of “veto” that coalition parties may have is if their votes are required for a bill to receive single majority support. Additionally, the Prime Minister has an instrument in their possession that is unique to them - the ability to at any point dissolve Parliament (Ibid., Section 57) - which could be used to get coalition partners or uncooperative factions of their own party to fall in line.

(3) There are no specific mechanisms in the constitution for guaranteeing proportional representation for different societal groups in regard to civil service appointments. The power to appoint high-ranking public servants, such as the Chief Justice, Ambassadors and High Commissioners abroad or members of the Public Service Commission, is vested in the Governor-General, acting in accordance with the advice of the Prime Minister (Ibid., Sections 77 & 87-88). This could theoretically be a hindrance to proportional representation, under the assumption that the Prime Minister may be more likely to advocate for the appointment of a person from his or her own community. The Public Service Commission, consisting of five members, is in turn responsible for appointing public servants in lower levels of government (Ibid., Sections 88-89), which may have further negative consequences for representativeness if this entity itself is not representative of the population. To balance this out, there is the

9 For the sake of simplicity, the number of members per constituency will henceforth be referred to as three, but the same logic applies to Rodrigues even if they only send two members.

office of the Ombudsman, which is meant to review and investigate complaints regarding the actions of members of the Assembly or other public servants. The Ombudsman is also appointed by the Governor-General, but acting in accordance with both the Prime Minister, the Leader of the Opposition and other party leaders (Ibid., Sections 96-97).

In regard to proportional representation in the Assembly, there is the Best Loser system, which provides additional seats in Parliament for underrepresented communities, from a roster of candidates from each of these groups who fell just short of winning a seat. It should be mentioned however that this system is limited to four groups - Hindus, Muslims, Chinese and the General Population; the last group encompasses all individuals who is not part of any of the other communities, for example both white Franco-Mauritians and the Creole population which mainly is of African descent (Ibid., Schedule 1).

(4) While there are no specific protections for groups to handle their own internal affairs, there are no passages that explicitly prevent such actions from being taken. The closest thing to such a section is one passage that grants all the various religious communities in Mauritius the right to independently provide education to members of their community, without the involvement of government or other religious groups (Ibid., Section 11); this can not be equated with full autonomy for these communities however.

On the whole, the constitution can not be said to contain many elements of consociationalism; the most notable exception is the Best Loser system for ensuring proportional representation. There are also some indications of autonomy and protection for the different communities, but whether this has any real significance will be analyzed further in a later section.

The second period (1992-2020) Since Mauritius constitution has remained largely the same since independence in 1968, this section will not go over all the questions once again, but will instead focus on what parts of the constitution has been amended since 1992, when the nation became a republic. The largest change from the first to the second period naturally is that the Head of State went from being a Governor-General acting on behalf of the British monarch to being a President elected by the National Assembly. The position remained primarily formal, however, with the main political power being vested in the National Assembly and the President assuming all the responsibilities that previously befell the Governor-General. Through an amendment in 2003, the powers of the presidential office were increased somewhat, with the president gaining the power to refuse a request from the Prime Minister to dissolve Parliament, as well as gaining additional powers to appoint certain public officials (Mathur, 2017, pp 373-376).

Another change that technically happened during the first period, but only came into effect during the second was an amendment in 1991, changing the Best Loser System so that the four best losers from the winning coalition no longer had to belong to a minority community

if no such candidates were available (Ibid., pp 99-100). Before this amendment, if no candidates existed which matched the criteria, the seat would just be left empty in order to not tip the balance further in favour of an already overrepresented group.

Fiji constitution

The first period (1970 - 1992) (1) The electoral system described in the 1970 constitution is a highly complex and unique system where the nation is divided into constituencies in five different ways, each of which returns a certain number of members for the House of Representatives (Fiji, 1970, Section 32). The first division creates twelve constituencies, each of which are meant to contain 1/12th of the total adult population that belong to the native population, iTaukei. The second does a similar twelve-way division, but instead based on the adult Indian population. The third separates the nation into three constituencies, each of which contains 1/3rd of the nation's remaining adult population, that is of neither Indian or iTaukei origin. The fourth creates ten constituencies that each contain 1/10th of the nation’s total adult population, and the fifth does the same but instead with only five constituencies. Each of these constituencies returns one member except for the ten-way division which returns two members for every constituency, for a total of 52 Parliamentarians. One underlying assumption in this system is that electors of different heritage will vote differently; this would in turn entail that there will be a plurality of parties in Parliament. If this assumption is accurate however is impossible to determine without looking at the empirical results from the elections, which will be discussed later. The constitution also stipulates that there will be a Senate, consisting of members appointed by the Governor-General on the recommendation of several actors, such as the Prime Minister, the Leader of the Opposition and the Great Council of Chiefs (Ibid., Section 45).

(2) As with Mauritius, there is no constitutional mechanism for granting veto to minority parties in a coalition. The main difference that merits mentioning is that in Fiji, any bill or law has to be passed in not only the HoR but also in the Senate, effectively granting the Senate a veto on decisions made by the HoR (Ibid., Section 53). This may grant an additional layer of protection for minorities, by preventing laws that benefit only one or a few groups, since the ethnic, religious and cultural composition of the Senate is at least partially determined separately from the composition of the HoR.

(3) As in Mauritius, the constitution of Fiji specifies a Public Service Commission appointed by the Governor-General, that in turn is supposed to handle all appointments to public offices, and the same issues that were discussed in regard to Mauritius applies here. However, the Fijian constitution includes a section that specifies that the Public Service Commission shall “ensure that, so far as possible, each community in Fiji receives fair treatment in the number and distribution of offices to which candidates of that community are appointed on entry” (Ibid., Section 104-105). Additionally, the office of the Ombudsman also exists in Fiji and functions largely the same as in Mauritius (Ibid., Section 112-113).

In the election to the HoR, voters are sorted into three different rolls; Fijians, Indians and voters that are neither Fijian nor Indian. 22 members of Parliament are elected from the Fijian electorate; 12 by voters registered on that roll and 10 by the national roll. The same applies for the Indian roll. 8 members are elected from the “General Electorate” that belongs to neither of the other groups; 3 by voters on that roll and 5 by the national electorate (Ibid., Section 32). This mechanism ensures that the two most prominent groups - ethnic Fijians and Indo-Fijians - will be represented in Parliament; it does however not guarantee that all ethnic groups will be represented, since the third group contains people from a variety of heritages, from Polynesian to European to Chinese, and there is no constitutional ground for ensuring that members from all these groups will be elected. Additionally, this system only takes into account ethnicity, and has no mechanisms for ensuring representation for religious or lingual 10 groups . Whether the constitution was successful at actually providing representation will be discussed later.

(4) Once again, there is no element of the constitution that explicitly prevents separate communities from handling their internal affairs, nor are there any sections that specifically protect this right. There is an almost identical section as that in Mauritius, which protects the rights of religious communities to provide independent education to their members (Ibid., Section 11). This protection is especially interesting in Fiji’s case, since the rest of the constitution places a heavy focus on protection for various ethnic groups while largely overlooking the religious divides.

While still lacking certain important elements, it could be argued that the electoral system for guaranteeing representation for the various ethnic groups alone is enough to make Fiji’s first constitution more consociational than that of Mauritius; it does nevertheless fall short of fully living up to the ideal type of a consociational democracy.

The second period (1992-2020) (1) The 2013 constitution of Fiji establishes that the nation’s electoral system will be proportionally representative, with the whole population being included in the same national roll and with seats in parliament being awarded to parties or independent candidates in proportion to the percentage of votes that they received in the election, under the condition that they pass the threshold of 5% of the total vote. In total, 50 members are elected for a single-house Parliament (Fiji, 2013, Section 53). Out of the three systems that have been discussed, this one should in theory provide the largest possibility for a plurality of parties; whether or not this is the case in practice remains to be seen.

10 This is arguably a larger issue for the Indian community than the Fijian, since the former is more internally heterogeneous both religiously and linguistically while the latter is more homogeneous with most people with iTaukei origin being Christian and speaking the Fijian language.

(2) As with the previous constitutions, there is no built-in veto for coalition parties in the constitution and the only person with any kind of veto is the President, who in this case is chosen from two candidates put forward by the Prime Minister and Leader of the Opposition respectively, which can not be members of or hold an office in any political party (Ibid., Sections 83-84). This independent status of the President could hypothetically offer some additional protection from bills or laws that attempt to infringe on the rights of any community.

(3) As for public servants, Fji’s 2013 constitution sets out, similarly to the other constitutions, that there will be a Public Service Commission that will be responsible for the recruitment of these individuals based on: “i. objectivity, impartiality and fair competition; and ii. ability, education, experience and other characteristics of merit” (Ibid., Section 123). These guidelines imply that ethnicity, religion or language should not be taken into consideration for the appointments to public offices, but simultaneously does nothing to ensure representation for these same offices. The jurisdiction of the Public Service Commission also does not extend to offices in the Fijian Military or Police Forces; nor does it extend to the appointment of judges, which is the responsibility of the Judicial Services Commission (Ibid., Sections 125-126).

The constitution further lacks any concrete mechanisms for ensuring political representation for the various communities in Fiji. The main guiding principle for the electoral system is proportional representation for parties or independent candidates; the constitution does not specify however whether there will be a possibility for voters to vote for a specific candidate within a party or if the lists of candidates are by the parties themselves. If the latter is the case, this may be detrimental to political representation since the power to determine the representation of different societal groups in Parliament is granted to the parties rather than the voters.

(4) The preamble separately recognizes the different ethnic groups in the nation - iTaukei, 11 Rotuman , Indians, Pacific Islanders and finally other immigrants and settlers - and declare that they are all part of the same people and the same nation. While this recognition of the different groups and their independent languages and traditions is symbolically important, there is little that separates this constitution from the 1970 one in regard to the actual ability for different groups to internally handle their affairs. One thing that can be mentioned is that this constitution explicitly specifies the secular nature of the state, which may make different religious communities more comfortable to practice however they please without the involvement of the state (Ibid., Section 4). Additionally, the constitution guarantees that iTaukei, Rotuman and Banaban lands will remain in the hands of their “customary owners'' and will neither be acquired nor sold by the state. This may be seen as one way to grant some degree of autonomy to certain groups, but it could also be seen as preferential treatment of

11 The Rotumans from Rotuma Island belong to a different ethnic group than the iTaukei, and in total make up approximately 1% of Fiji’s total population (WPR, 2020a)

indigeneous groups at the expense of communities with a different heritage (Ibid., Section 28-29).

As has been hinted at, the constitution has an aura of attempting to bridge communal divides and instead promote coexistence and cooperation between all the societal sections, but it does little to actually ensure that this will be the outcome.

Mauritius in practice12

The first period (1968-1992) (5) Mauritius has a long history of political parties forming electoral alliances which later turn into broad coalitions in the event that the alliance wins the election. This can be traced back to even before independence; in the 1967 election the Labour Party, the Independent Forward Bloc (IFB) and the Muslim Action Committee (CAM) formed an alliance headed by the leader of the Labour Party, Sir Seewoosagur Ramgoolam (Mathur, 2017, pp 57-60; Bowman, 1991, pp 37-42). This alliance - known as the Independence Party - went on to win the 1967 election and would lead the country through its first 8 years of independence. While the coalition did not represent all segments of Mauritian society, it did contain several of the most important groups; the Labour Party was primarily made up of a mix of Hindus and Creoles, while the IFB and CAM were dominated by Hindus and Muslims respectively (Bowman, 1991, pp 36). The next election in 1976 would result in yet another broad coalition government, this time between the Labour Party, CAM and the Mauritian Social Democratic 13 Party (PMSD), one of two parties representing the General Population - thus adding Franco-Mauritians to the represented groups (Mathur, 2017, pp 60-63). The only major group still lacking representation was Sino-Mauritians; however, many people belonging to this community identified with the politics of the PMSD (Bowman, 1991, pp 69-70).

Four more elections were held during the last 10 years of this time period - in 1982, 1983, 1987 and 1991 - all resulting in coalition governments led by the same Prime Minister: Anerood Jugnauth. In 1982 he was elected as a member of the MMM - which came to power in coalition with the Mauritian Socialist Party (PSM) - but from 1983 and forward he served as leader of the newly formed Militant Socialist Movement (MSM) (Ibid., pp 82 - 102; Mathur, 2017, pp 63-72). Between 1983 and 1991 the MSM built an alliance with the Labour Party, notably joining together the two largest parties with a voter base primarily consisting of Hindus; this was overturned in 1991 however when the MSM coupled with the MMM, once again creating a coalition representing several of the major ethno-religious communities (Ibid.).

12 For a complete list of Mauritian leaders over the years, see Appendix 13 The other being the Mauritian Militant Movement (MMM) - founded and headed by Paul Bérenger - which despite winning the most seats of any party (30) was left outside of the coalition (Mathur, 2017, pp 61).

(6) As was mentioned during the section on previous research, Carroll and Carroll in an article from 2000 (pp 120-142) found that civic society played an important role on the top-level of Mauritian politics during the period from 1980-1995, with unions, corporations and organization representing essentially all subsections of society being consulted by the government on important political issues. This practice began in the late 1970’s when the nation first started developing its National Plans, and has since been used frequently for several different areas of politics. In interviews with representatives for various organizations the authors further found that this was not simply a symbolic gesture but that the consulted parties actually perceived that the government was listening to them and heeding their advice (Ibid.). This aura of inter-group cooperation can be traced back even further, to the pre-independence days of consultations regarding the design of the 1947 and 1968 constitutions (Bowman, 1991, pp 32-33, 37-42).

(7) As the section on constitutional conditions found, there is no formal veto mechanism for parties in a coalition to block decisions that they believe to be harmful to the interests of the communities that they represent; however, there may be other ways in which parties can influence the decisions of their governing partners. One example would be by leaving or threatening to leave the government, though this presupposes that the votes of the party in question are needed for the Prime Minister to have the support of a majority of the Parliamentarians. There is one occurrence during the researched time period where this actually took place - in 1983 - when Prime Minister Jugnauth split from the MMM after a falling out with the party leader and Finance Minister, Bérenger. This split was in part caused by Jugnauth - a member of the Hindu community - perceiving that some of Bérengers policies favoured Franco-Mauritians and Creoles, resulting in him leaving the party and forming the MSM, which almost immediately absorbed the PSM (Ibid., pp 82-88). While this is clearly not as effective nor practical as a formal veto, it still shows that there were other methods which parties or individuals could, and in fact did use in order to oppose policies they perceived as damaging to their electorate.

(8) Since the main instrument for ensuring political representation in Mauritius is the Best Loser System, this section will analyze its effectiveness. Mathur (1997, pp 61-82) exemplifies how the BLS works by looking at the 1987 election and how the eight best loser seats were distributed. The original election results provided the most seats for candidates belonging to the Hindu community, followed by the General Population, Muslim community and finally the Chinese community (see Table 2). Based on this result, Mathur calculates the quotient of people from each respective community divided by the seats won by candidates from that community plus one, showing that the most underrepresented segment was the Muslim community; this process was then repeated for all eight seats, which in this case resulted in the General Population and Muslim community each being awarded four seats.

Table 2 Hindu General Muslim Chinese Population

Number of seats 40 (10 448) 15 (16 340) 6 (19 596) 1 (12 187) before Best Loser (quotient)14 Number of seats after 40 (10 448) 19 (13 072) 10 (13 717) 1 (12 187) Best Loser (quotient) Adapted from Mathur (1997, pp 64-70) Data from 1972 Population Census of Mauritius & Government Gazette Extraordinary (1987)

While the Hindu population was still somewhat overrepresented, the distribution is considerably more balanced than before, which shows that the BLS does in fact work in practice, though with some limitations. As for other public offices, in accordance with what was argued in a previous section, it is reasonable to assume that proportional representation in the National Assembly will also result in representation in the lower levels of the public sector.

(9) One significant divide in Mauritian society during the time period regards the different language groups. The three largest language groups in the country in 1983 were Kreol, Bhojpuri and Hindi, being the first spoken languages of 54%, 20% and 11% of the population respectively (Ibid., pp 53). In the educational system, however, English is the lingua franca 15 on all levels and from Standard 3 and up all students are also required to learn French (Ibid., 56). English and French are also the two languages used in the National Assembly (Mauritius, 1968, Section 49). The use of these two languages, spoken by 0,21% and 3,7% of the population respectively, instead of granting special rights to any of the larger languages has avoided conflict over which language should be dominant. The “traditional” languages of the various communities are primarily promoted by the religious institutions (Eisenlohr, 2018, pp 191-211 in Ramthoul & Eriksen), granting these an additional level of autonomy.

It can further be observed that Mauritian public holidays draw from a wide variety of different cultural and religious practices and traditions, primarily Christian, Hindu, Muslim and Chinese (Bowman, 1991, pp 51); this serves both to give each religious segment the freedom to practice their traditions while also bridging the gaps between the different segments.

14 This quotient shows how many people from that community that each member would be representing if they were granted one additional seat. 15 In lower standards Kreol and Bhojpuri are used until students have a sufficient comprehension of English (Bowman, 1991, pp 56)

While certain elements - such as a mutual veto for coalition parties and a more extensive autonomy for the various segments - are still lacking for Mauritius to fully live up to the ideal type of a consociational democracy, many of the elements that Lijphart describes can indeed be found in the nation during the period.

The second period (1992-2020) (5) The tradition of coalition governments with broad representation has continued during the second period, which to a large part has been dominated by the same three parties as the first - MSM, MMM and Labour. The 1995 election saw a total victory for a new coalition between the MMM and Labour, which together won all the 60 contested seats (Mathur, pp 72-74). The Prime Minister post was granted to Navin Ramgoolam, the son of Seewoosagur, who had assumed the leadership position of the Labour Party in 1991, following his father’s death in 1984 (Ibid. pp 70). This coalition only lasted until the next election in 2000, when the MSM and the MMM once again allied and won the election; this new alliance was built on a power-sharing arrangement where Jugnauth would be Prime Minister for the first three years 16 with Bérenger as Deputy, followed by two years with Bérenger as PM and Jugnauth’s son, Pravind, assuming the position of Deputy (Ibid., pp 74-76). Jugnauth Sr. instead became President of the Republic, a position which he held between 2003 and 2012 (Lansford, 2017b, pp 982). The 2005 election yet again saw the Labour Party come into power, as the leader of the Social Alliance - a broad coalition consisting of five parties with various political identities (Ibid., pp 76-78). This victory would be followed by yet another one in 2010, this time for a coalition composed of the Labour Party, the MSM and the PMSD (IFES, 2010). The final two elections of this period, in 2014 and 2019, were both won by an alliance between the MSM and the PMSD (IFES, 2014b; IFES, 2019); this coalition was at first led by Anerood Jugnauth, but in 2017 he stepped down and passed over the Prime Minister post to his son, Pravind (WPR, 2017). While this alliance represents the two largest groups in Mauritius - Hindus and the General Population - notably lacking during the entire period is any party specifically representing the interests of either the Muslim or the Chinese community.

(6) The findings by Carroll and Carroll discussed above also extend to the first years of this period, though the authors note that on one important decision in 1994/95 the government did not consult with the civic networks; this was the proposed policy to allow communal languages to gain the same educational status as English and French and thus be allowed in 17 examinations for secondary school applications (Carroll & Carroll, 2000, pp 131-132). In the economic field there has been a similar involvement from other actors, mainly through 18 the Joint Economic Committee (JEC) which was established all the way back in 1970 to 16 This marked the first and so far only time in Mauritian history when the Prime Minister position has been held by a non-Hindu and a person not belonging to the Ramgoolam or Jugnauth families. 17 This proposal was never implemented since the MSM lost all their seats in the 1995 election. 18 In 2015, the JEC merged with another organization, the Mauritius Employers Federation, in order to create the new organization Business Mauritius, which works largely the same as the JEC did previously.

promote dialogue between the government and the private sector (Darga, 2020, pp 40-47, in Seetanah et al.). While the JEC encompasses a broad spectrum of firms and corporations, it lacks a mechanism for ensuring that the interests of all communities are represented in the conversation.

(7) As has been made clear by now, there is no formal veto mechanism by which parties can block government policies; there has however been instances where other actions have been taken to oppose government policies. One such action took place in 2013, when Prime Minister Ramgoolam declared his support for proposed changes to the electoral system, such as removing what community a candidate belongs to from the voter ballots, and reducing the number of constituencies in order to limit the number of parties and consequently reducing the need for coalition governments (Lansford, 2017b, pp 982-983). These proposals were rejected by Vice Prime Minister Xavier-Luc Duval - the leader of second-largest government party PMSD - who resigned from his post in June 2014 and withdrew his party from the 19 coalition (Ibid.) . The proposals were never passed into legislation, and in December the same year the Labour Party lost the election to a new coalition where the PMSD was included; it can thus be argued that Duval in effect stopped the legislation when he resigned.

(8) The most notable change regarding proportional representation from the first to the second period was, as stated above, that the Best Loser seats for the winning coalition from the 1995 election and onward no longer had to be given to under-represented communities, if no candidates were available. It is therefore interesting to see whether this has impacted representation in government. Mathur (2017, pp 89-96) analyzes the allocation of seats after the 2005 election in the same way as he did for the 1987 election, and finds that in this election the General Population was most underrepresented before the Best Loser System came into play, followed by the Muslim community (see Table 3). Through the BLS the General Population was thus awarded 6 seats and the Muslim community 2, making the final allocation identical to that of 1987.

Table 3 Hindu General Muslim Chinese Population

Number of seats 40 (10 448) 13 (18 674) 8 (15 241) 1 (12 187) before Best Loser (quotient)

Number of seats after 40 (10 448) 19 (13 072) 10 (13 717) 1 (12 187) Best Loser (quotient) Adapted from Mathur (2017, pp 89-96) Data from 1972 Population Census of Mauritius

19 This left the Labour party as the sole governing party, since the MSM had previously pulled out of the coalition in 2011 following the arrest of Health Minister Hanoomanjee on charges of corruption (Ibid.)

While the 2005 election saw no distortion due to the 1991 amendment, this did happen in the 2000 election, when two seats intended for the General Population instead went to Hindu candidates, thus skewing the representation even further in favour of the group that already was the most overrepresented (Mathur, 2017, pp 99-100). This may be an issue for proportional representation during the entire period, since the BLS no longer is a guarantee for the traditionally underrepresented communities to get additional representation in the parliament.

(9) No significant changes have taken place in this area from the first to the second period; public holidays still include important celebrations for a variety of the most important communal groups (GovMU, 2019) and there is still no official national language, but the languages used for official purposes and within education are still English and French (Nag, 2017).

On the whole, Mauritius in the second period very much resembles Mauritius in the first period, with a continued tradition of broad coalitions representing the different sections of society, but still lacking certain elements to make it fully consociational.

Fiji in practice20

The first period (1970-1992) (5) The first 17 years of Fiji’s independent political life were dominated by one man and one 21 party - Ratu Sir Kamisese Mara, leader of the Fijian Alliance Party (FAP) - which first came into power after the election in 1966 (Lansford, 2017a, pp 499-500). While the FAP was composed ethnically of candidates from several different societal groups, its support overwhelmingly came from the ethnically Fijian electorate (Hartmann, 2001, pp 644). The main contender to the FAP during this period was the National Federation Party (NFP) - led by Siddiq Koya - which mainly drew its support from the Indo-Fijian group. In the years right around independence, there was significant collaboration between these two parties on constitutional and electoral issues and there was even a proposal in 1971 for them to form a coalition which would have been a significant step toward consociationalism (Howard, 1991, pp 78). After the 1977 election this proposal came up again since neither party managed to gain a majority of seats but was once again turned down in favour of a re-election (Ibid, 1991, pp 97-99). The second election of 1977 saw a conclusive victory for the FAP, as did the 1982 election when the party once again won a majority without needing to form a coalition (Hartmann, 2001, pp 666). 1987, however, saw the first electoral victory for a coalition in Fijian history, when the newly formed Labour Party (FLP) in coalition with the NFP won the election with 28 seats to the 24 seats of the FAP (Ibid.). The coalition was led

20 For a complete list of Fijian leaders over the years, see Appendix 21 Ratu is a hereditary chiefly title, owing to the fact that Mara was paramount chief of the Lau archipelago in eastern Fiji (Economist, 2004, pp 100).

by Timoci Bavadra, a member of the iTaukei community, but mainly drew its support from Indo-Fijians and working-class people from the indigenous Fijian community (Ibid., pp 645).

(6) In the early years of Fiji’s independence, trade and labour unions had a significant influence on politics, and the Fiji Trades Union Congress (FTUC) had a close cooperation with the Alliance Party (Howard, 1991, pp 85). Eventually this collaboration developed into a more institutionalised partnership through the establishment of the Tripartite Forum in 1977, through which representatives for the labourers, the employers and the government could meet to negotiate wages, workplace conditions and matters of economic policy (Ibid., pp 105-106). While initially rather successful, the cooperation met a hurdle in 1985 when the government decided to drive through legislation that was contrary to agreements made through the Forum and that benefitted employers at the cost of labourers; this was one contributing factor to the formation and eventual success of the FLP in the 1987 election (Ibid., pp 151-153). It should also be noted that the Forum had no mechanism for ensuring communal representation, thus limiting its consociational function.

(7) As Fiji’s sole coalition government during the period only lasted a month before being overthrown in the first of two military coups in 1987, there is little precedent to use for determining the existence of an informal veto, but as with Mauritius there are a few instances of other “protest actions”. One example came from the Indian Alliance - a wing of the FAP representing Indo-Fijian interests. Beginning in the late 1970’s, accusations started arising against Mara of being negligent regarding the interests of Indo-Fijians, eventually culminating in the president of the Indian Alliance, James Shankar Singh, leaving the party in protest in 1982 (Ibid., pp 109). This action was symbolically important but did not produce any effect; the FAP did not alter their policies and went on to win the election later in the year.

(8) As was mentioned above, the 1970 constitution set out a very complex system for ensuring representation for the three major ethnic groups by providing each with a number of parliamentarians proportional to their share of the population. The number of parliamentarians from each community remained the same for the entire time period - the Indo-Fijians and iTaukei got 22 seats each while the General Electorate got 8. What needs examining is thus whether the relative sizes of the various voter groups have changed over time. Table 4 shows how the number of registered voters from each group has changed between elections and also how their quota of the total electorate has changed.

Table 4 iTaukei Indo-Fijian General Electorate Total (quota) (quota) (quota)

1972 98 192 (0,478) 100 846 (0,49) 6 579 (0,032) 205 617

1977(Apr) 137 940 (0,491) 134 798 (0,48) 8 046 (0,029) 280 784

1977(Sep) 140 430 (0,489) 137 968 (0,481) 8 683 (0,030) 287 081

1982 141 846 (0,485) 142 529 (0,488) 7 966 (0,027) 292 341

1987 169 398 (0,479) 174 611 (0,494) 9 682 (0,027) 353 691 Data from Hartmann (2001, pp 660-665)

As can be seen, the relative sizes of the groups have stayed largely the same over time, even if the General Electorate has grown at a somewhat slower pace than the other two groups. What is more noticeable, however, is the fact that the General Electorate holds about 15% of the seats in Parliament despite only comprising roughly 3% of the total electorate, thus being massively overrepresented in comparison to the other two groups. Lastly, it merits mentioning that the 1990 constitution completely did away with the old distribution system in favour of one weighted heavily in favour of the iTaukei; this did not affect the distribution of seats until the election of 1992 however and will thus be discussed in relation to the next period.

(9) As has been alluded to several times, the political sphere during this period was dominated by the indigenous Fijian community, with only limited possibilities for people from other communities to gain any real influence. From this, it can be extrapolated that the iTaukei were relatively free to make autonomous decisions while the other communities were not. This tradition of dominance was according to Howard (1991, pp 9-11), one reason why many ethnic Fijians saw the 1987 victory of the FLP-NFP coalition as threatening to their interests, which led to massive protests and eventually a military takeover.

Fiji’s democratic success for most of this period was very much dependent on the dominant position of the Alliance Party and the iTaukei community, and even though there were stretches of inter-ethnic cooperation, these did not develop into a more institutionalized consociational system.

The second period (1992-2020) (5) While the more recent period of Fiji’s history has been considerably more tumultuous than the former, there have also been periods of relative stability and throughout the period, seven elections have been held. The first two elections in 1992 and 1994, saw the newly formed party Soqosoqo ni Vakavulewa ni Taukei (SVT) - led by the military leader of the 1987 coups, Sitiveni Rabuka - winning the most seats in the parliament (Hartmann, 1991, pp 666). While not strictly in coalition with each other, his premiership hinged on the votes of the FLP, which in exchange demanded constitutional reform (Ibid., pp 646). The new 22 constitution of 1997 opened for the FLP to win the election in 1999 , forming a coalition with the Party of National Unity and the Fiji Association Party - two parties primarily

22 One feature of the constitution was that any party with more than 10% of the seats could take a place in the government (Hartmann, 2001, pp 646)

supported by indigenous Fijians (Ibid.). Together, the coalition held 51 out of the 70 seats in Parliament and represented the interests of several ethnic groups, making it the closest example in Fiji’s history of a “broad coalition”. The leader of FLP, Mahendra Chaudry, was appointed Prime Minister, thus making him the first and so far only Indo-Fijian to hold the position; he would only retain it for roughly a year however before the government was overthrown in a civilian coup (Ibid.). After a brief period 23 of military rule, Laisenia Qarase was appointed as interim Prime Minister and a new election was scheduled for 2001. Qarase’s newly formed party Soqosoqo Duavata ni Lewenivanua (SDL) went on to win the elections of both 2001 and 2006 only to be overthrown in a military coup in 2006, led by Commodore Frank Bainimarama (McCarthy, 2011, pp 566). Bainimarama installed himself as interim Prime Minister and the military effectively continued to rule the country until 2014, when the next election was held (Lansford, 2017a, pp 500). Neither of the final two elections of the period saw any coalition governments, as FijiFirst - headed by Bainimarama - won a majority of the seats in Parliament in both 2014 and 2018 (IFES, 2014a; IFES, 2018).

(6) While there has not been any institutions during the period with the function of counseling the government, non-governmental committees have still played important roles on at least two occasions - when drafting the constitutions of 1997 and 2013. The 1997 constitution came to be after a Constitutional Review Commission was set up in 1995 in order to propose changes to the much criticized - both domestically and internationally - 1990 constitution (Hartmann, 2001, pp 646). The commission consisted of three members, Tomasi Rayalu Vakatora, a former senator and Speaker of the house of iTaukei origin, Brij V. Lal, an Indo-Fijian scholar specialising in Fiji’s political history, and Sir Paul Reeves, former Governor-General of New Zealand of European and Māori origin (Lal, 1997, pp 71; GovNZ, 2020). This commission thus represented three of the largest and most important groups in Fiji. Another commission was appointed in 2012, headed by Kenyan scholar Yash Pal Ghai, in order to draft a new constitution that was to be in place by 2013 (Lansford, 2017a, pp 501). Following an original draft by Ghai, however, the government took control of the process and made significant changes to the document (ABCNews, 2013), thus essentially eliminating the purpose of an independent commission.

(7) As with the first period, there is only one relatively short period of an actual “broad coalition” and thus there is not much data to analyze. Nevertheless, there is one action during the period that still merits mentioning; the 1998 split within the SVT. This event was caused by dissent within the party following the 1997 constitution and continued talks between the government and opposition, which led some of the party’s indigenous members to split and form the Veitokani ni Lewenivanua Vakarisito (Bissessar, 2017, pp 33). In the following election in 1999, that party won 19,4% of the iTaukei vote and 9,8% of the Open vote, and

23 Qarase was of ethnic Fijian descent and had previously been the head of the Fiji Development Bank (Lal, 2019, pp 323).

24 may thus have been one contributing factor to the SVT losing power (Hartmann, 2001, pp 660-665).

(8) For the first two elections of the period, in 1992 and 1994, seats in parliament were distributed according to the 1990 constitution, with 37 seats going to the iTaukei, 27 to Indo-Fijians, 5 to the General Electors and 1 to the Rotumans, for a total of 70 seats (Hartmann, 2001, pp 649). The number of registered voters for each of these elections can be seen in Table 5.

Table 5 iTaukei Indo-Fijian General Electorate25 Total (quota) (quota) (quota)

1992 154 004 (0,492) 148 548 (0,474) 10 642 (0,034) 313 194

1994 155 681 (0,472) 159 480 (0,483) 14 931 (0,045) 330 092 Data from Hartmann (2001, pp 660-665)

As can be seen, the balance between the different groups stayed largely the same as in the first period; however, since the iTaukei now were guaranteed 37 seats (~53%) while the Indo-Fijians only received 27 (~39%), the former group was now slightly overrepresented while the latter was quite seriously underrepresented. In regard to the General Electorate, they were still overrepresented with 5 seats (~7%) but not as severely as before.

The constitutional amendments in 1997 altered the distribution to once again be more proportional, with 23 seats for the iTaukei, 19 for Indo-Fijians, 3 for General voters and 1 for Rotumans. Further, an Open Roll was established for all voters, regardless of ethnicity, which elected an additional 25 members (Ibid., pp 650-651); unfortunately this also entailed that there no longer was any statistics on the communal belonging of the elected candidates. To substitute for this, it can be noted that out of the three parties that made it into Parliament in 2014 and 2018, the two major ones - FijiFirst and Social Democratic Liberal Party (SODELPA) - had leaders from the iTaukei community while the smaller one - NFP - had an 26 Indo-Fijian leader (IFES 2014a; IFES 2018) .

(9) While this period has seen many significant changes in the political structure of the nation, not a lot has changed in regard to this area. One thing that can be mentioned is that the 2013 constitution sets out that all school education shall take place in English and that all primary school children have to learn both iTaukei and Fiji Hindi at a basic level (Bissessar,

24 The SVT went from winning 32 seats (45,7% of the vote) in the 1994 election to only winning 8 seats (11,3% of the vote) in 1999. 25 For the 1994 result, Rotumans are included in the General Electorate while for the 1992 result they are left out since their seat was uncontested; this explains the relatively big increase in General Voters from 1992 to 1994. 26 Leader for FijiFirst was Prime Minister Bainimarama, while the NFP was led by Biman Prasad. SODELPA changed its leader from Ro Teimumu Kepa in 2014 to Sitiveni Rabuka in 2018.

2017, pp 38); this compulsory education may serve to bridge the gaps between communities, but may simultaneously limit their individual autonomy.

With the exception of the relatively short interval from the formation of the Constitutional Commission in 1995 to the coup in 2000 there are not many examples of consociational elements in the Fijian system during this period; that interval, however, is doubtlessly the closest Fiji has been to a consociational democracy during its entire 50-year existence.

Results and discussion

Analysis summary and interpretation Weighing together both the constitutional and practical results, Mauritius in both the first and second period exhibits clear traits associated with consociational democracy; a tradition of broad coalitions, a system for making political representation more proportional and some degree of autonomy for various ethnic, religious and lingual groups. Certain elements are still missing for it to fully live up to the ideal type based on Lijphart’s four criteria - notably a mutual veto for all parties in a coalition. As has been shown, however, some actors have found other ways to effectively block legislation or decisions that they believe would be damaging to the interests of the group that they represent. One remarkable feature of the Mauritian system is that many of the institutions that today make it consociational have not been engineered as part of the constitution but have rather developed naturally through the actions of politicians and members of civic society; even the Best Loser-system originally came to be as a result of political compromise during the constitutional talks in the 1960’s (Bowman, 1991, pp 40).

In Fiji, the first period does show a few signs of consociationalism in both the constitutional and practical field - the system for proportional representation stands out as the most clear example - but compared with Mauritius and with the ideal type, there are still many features lacking. Worthy of note, though, is that Fiji was still clearly democratic for most of this period, until 1987, even if political life was mainly dominated by the FAP. The second period has seen several attempts at bringing the country back in a more democratic direction, most notably with the 1997 constitution which had a few consociational traits. These attempts have mostly been unsuccessful, however, which in part may be attributed to the fact that they were engineered rather than naturally evolved, as in Mauritius.

In regard to the time dimension, one thing to mention in Mauritius is the early seeds of inter-group cooperation that began already in the period leading up to independence, only to then continue to grow even stronger in the years that followed thereafter. The participation of all sections of society during this crucial period laid the foundation for continued collaboration even in the later period, and has been essential in ensuring Mauritius political

and social stability over time. This stability in turn has allowed democracy to not only evolve but flourish. In Fiji the analysis provides a different result; while the days just before and after independence saw cooperation between groups, the decision by the FAP in the 1970’s to turn down a coalition government meant the practice was not institutionalized in the same way that it was in Mauritius, and political power was at the outset largely reserved for indigenous Fijians. When this hegemony eventually was disturbed in 1987, conflict could no longer be contained within the field of politics and the military intervened; this pattern was then repeated several times during the second period and became the main obstacle for political and social stability in Fiji. The effects on democracy have been detrimental, with any progress being quite rapidly followed by new setbacks.

Other possible explaining variables Carroll’s article on a similar topic that was mentioned in a previous chapter found that the main difference between Fiji and Mauritius is that Fiji’s indigenous population - the iTaukei - claim a special status as the original inhabitants of the islands while no group in Mauritius make a similar claim. This finding is not incompatible with the results of this paper; rather, Carrol’s finding fits in neatly as part of the explanation proposed herein. If the iTaukei claim a special political status, this would explain the reluctance of the FAP - and later the SVT, SDL and FijiFirst - to cooperate and form coalitions across communal lines, instead preferring to remain the sole governing parties. Further, if no group in Mauritius claims such a status, this explains why coalition partners have been able to see each other as equals, thus facilitating cooperation rather than prohibiting it.

Future research Continued research on this subject may entail expanding the comparison to include more countries that resemble Fiji and Mauritius in terms of size, plural society, economic preconditions and colonial heritage, but that has different variations on the dependent variable - level of democracy; a few examples of such nations may include Trinidad & 27 Tobago, Guyana and Suriname . Such a study would take the same form as this one, by attempting to determine the degree of consociationalism in order to find out if this can explain the variation in democracy, thus expanding the generalizability of the hypothesis.

Since this paper has shown that Mauritius has more consociational elements than Fiji, the next logical question is what has caused this system to develop in one nation and not the other. Another study could therefore potentially look at another concept retrieved from Lijphart, that was only briefly mentioned in this study, namely favourable and unfavourable conditions for consociational democracy (Lijphart, 1977, pp 53-103). This would be another link in the longer chain of attempting to discern which factors that are relevant for the development of democracy.

27 Suriname has a Dutch colonial heritage, but in all other regards very much resembles the others.

Conclusion The goal of this paper was to find the answer to the question: “Why has Mauritius developed a functioning democracy when Fiji has failed to do so?” by using Arend Lijphart’s theory of consociational democracy. Through comparing the two nations, and by adding the dimension of time in order to determine cause and effect, it has become apparent that the existence - or lack - of consociational elements in the early days of independence was a critical component of Mauritius success and Fiji’s failure to evolve and sustain democratic rule. With the way the study was designed - using a Most Similar Systems Design - these results may also be able to say something more generally about why some plural, post-colonial societies democratize while others do not. To properly determine their applicability, however, further research is needed.

References

ABCNews, 2013, Constitutional expert Yash Ghai says he doubts if Fiji PM Frank Bainimarama has read constitution, Australian Broadcasting Corporation, October 23, available at https://www.abc.net.au/news/2013-10-23/an-constitution-expert-yash-ghai-says-he-doubts-if- fiji-pm-has-/5039704 [Accessed December 29, 2020]

Bissessar, A.M., 2017, Ethnic Conflict in Developing Societies: Trinidad and Tobago, Guyana, Fiji, and Suriname, Palgrave Macmillan, New York

Bowman, L.W., 1991, Mauritius: Democracy and development in the Indian Ocean, Westview Press, Boulder

Carroll, B.W., & Carroll T., 1999, The consolidation of democracy in Mauritius, pp 179-197 in Democratization, 6(1)

Carroll, B.W., & Carroll, T., 2000, Accommodating ethnic diversity in a modernizing democratic state: theory and practice in the case of Mauritius, pp 120-142 in Ethnic and Racial Studies, 23(1)

Carroll, T., 1994, Owners, immigrants and ethnic conflict in Fiji and Mauritius, pp 301-324 in Ethnic and Racial Studies, 17(2) Bräutigam, D., 1997, Institutions, Economic Reform, and Democratic Consolidation in Mauritius, pp 45-62 in Comparative Politics, 30(1)

Chimakonam, J., Agu, J., & Agbo, S., 2014, The Nemesis of Individualistic Ontology in Globalization and the Practice of Liberal Democracy in Post-Colonial Africa, pp 137-155 in The Journal of Pan African Studies, 7(4)

CIA, 2020a, The World Factbook: Australia-Oceania, Fiji, Central Intelligence Agency, available at https://www.cia.gov/library/publications/the-world-factbook/geos/fj.html#People [Accessed December 29, 2020]

CIA, 2020b, The World Factbook: Africa, Mauritius, Central Intelligence Agency, available at https://www.cia.gov/library/publications/the-world-factbook/geos/mp.html#People [Accessed December 29, 2020]

Economist, 2004, Obituary: Kamisese Mara, pp 100 in The Economist, 24 April, retrieved from The Economist Historical Archive, 1843-2015

EIU, 2020, Democracy Index 2019, The Economist Intelligence Unit, available at

https://www.eiu.com/public/topical_report.aspx?campaignid=democracyindex2019 [Accessed December 29, 2020]

Fiji, 1970, The Fiji Independence Order 1970, retrieved from HeinOnline, World Constitutions Illustrated

Fiji, 2013, Constitution of the Republic of Fiji, retrieved from HeinOnline, World Constitutions Illustrated

Freedom House, 2020, Countries and Territories, Freedom House, available at https://freedomhouse.org/countries/freedom-world/scores [Accessed December 29, 2020]

GovMU, 2019, Republic of Mauritius - Public Holidays 2020, Prime Minister’s Office, available at https://pmo.govmu.org/Communique/Public%20holidays%20--%202020.pdf [Accessed December 29, 2020]

GovNZ, 2020, Paul Reeves Biography, New Zealand History, available at https://nzhistory.govt.nz/people/paul-reeves [Accessed December 29, 2020]

Hartmann, C., 2001, Fiji Islands, in Nohlen, D., Grotz, F., & Hartmann, C., Elections in Asia and the Pacific: A Data Handbook: Volume II: South East Asia, East Asia, and the South Pacific, Oxford University Press, Oxford

Howard, M.C., 1991, Fiji: Race and Politics in an Island State, UBC Press, Vancouver

Huntington, S.P., 1991, The Third Wave: Democratization in the Late Twentieth Century, University of Oklahoma Press, Norman

IFES, 2010, Republic of Mauritius - ELECTION FOR ASSEMBLÉE NATIONALE (MAURITIAN NATIONAL ASSEMBLY), ElectionGuide, available at https://www.electionguide.org/elections/id/1560/ [Accessed December 29, 2020]

IFES, 2014a, Republic of Fiji - ELECTION FOR FIJIAN HOUSE OF REPRESENTATIVES, ElectionGuide, available at https://www.electionguide.org/elections/id/2564/ [Accessed December 29, 2020]

IFES, 2014b, Republic of Mauritius - ELECTION FOR ASSEMBLÉE NATIONALE (MAURITIAN NATIONAL ASSEMBLY), ElectionGuide, available at https://www.electionguide.org/elections/id/2822/ [Accessed December 29, 2020]

IFES, 2018, Republic of Fiji - ELECTION FOR FIJIAN PARLIAMENT, ElectionGuide, available at https://www.electionguide.org/elections/id/3131/ [Accessed December 29, 2020]

IFES, 2019, Republic of Mauritius - ELECTION FOR ASSEMBLÉE NATIONALE (MAURITIAN NATIONAL ASSEMBLY), ElectionGuide, available at https://www.electionguide.org/elections/id/3264/ [Accessed December 29, 2020]

Iyer, V., 2007, Enforced consociationalism and deeply divided societies: some reflections on recent developments in Fiji, pp 127-153 in International Journal of Law in Context, 3(2)

Lal, B.V., 1997, Towards a united future: Report of the Fiji constitution review commission, pp 71-84 in The Journal of Pacific History, 32(1)

Lal, B.V., 2019, Laisenia Qarase’s missed chance - In George Speight’s shadow, pp 323-348 in Levelling Wind: Remembering Fiji, ANU Press, Acton

Lansford, T., 2017a, Fiji, pp 500-585 in Political Handbook of the World 2016-2017, CQ Press, London

Lansford, T., 2017b, Mauritius, pp 980-986 in Political Handbook of the World 2016-2017, CQ Press, London

Lawson, S., 2012, Indigenous Nationalism, “Ethnic Democracy,” and the Prospects for a Liberal Constitutional Order in Fiji, pp 293-315 in Nationalism and Ethnic Politics, 18(3)

Lijphart, A., 1977, Democracy in Plural Societies, Yale University Press, New Haven

Lipset, S.M., 1960, Political man: the social bases of politics, Doubleday, New York

Mathur, H., 1997, Parliamentary Representation of Minority Communities: The Mauritian Experience, pp 61-82 in Africa Today, 44(1)

Mathur, H., 2017, Parliament in Mauritius, 2nd edition, Editions de l'Ocean Indien, Rose-Hill

Mauritius, 1968, The Mauritius Independence Order 1968, retrieved from HeinOnline, World Constitutions Illustrated

McCarthy, S., 2011, Soldiers, chiefs and church: unstable democracy in Fiji, pp 563-578 in International Political Science Review, 32(5)

Nag, O.S., 2017, What Languages Are Spoken In Mauritius?, WorldAtlas, August 1, available at https://www.worldatlas.com/articles/what-languages-are-spoken-in-mauritius.html [Accessed December 29, 2020]

Ramtohul, R., & Eriksen, T.H., 2018, The Mauritian paradox: fifty years of development, diversity and democracy, University of Mauritius, Réduit

Seetanah, B., Sannassee R.V., & Nunko, R., 2020, Mauritius; A successful small island developing state, Routledge, New York

UNDP, 2019, Human Development Report 2019, United Nations Development Programme, available at http://hdr.undp.org/en/content/human-development-report-2019 [Accessed December 29, 2020]

Witsoe, J., 2011, Rethinking Postcolonial Democracy: An Examination of the Politics of Lower-Caste Empowerment in North India, pp 619-631 in American Anthropologist, 113(4)

World Bank, 2020, GDP per capita (current US$) - Fiji, Mauritius, available at https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?locations=FJ-MU&most_recent_va lue_desc=true [Accessed December 29, 2020]

WPR, 2017, Mauritius’ Democracy Tarnished After Prime Minister Passes Power to Son, WorldPoliticsReview, January 31, available at https://www.worldpoliticsreview.com/trend-lines/21053/mauritius-democracy-tarnished-after -prime-minister-passes-power-to-son [Accessed December 29, 2020]

WPR, 2020a, Fiji Population 2020, World Population Review, available at https://worldpopulationreview.com/countries/fiji-population/ [Accessed December 29, 2020]

WPR, 2020b, Mauritius Population 2020, World Population Review, available at https://worldpopulationreview.com/countries/mauritius-population/ [Accessed December 29, 2020]

Appendix

Mauritius Prime Ministers 1967 - 1982 Sir Seewoosagur Ramgoolam 1982 - 199528 Sir Anerood Jugnauth 1995 - 2000 Dr. Navin Ramgoolam 2000 - 2003 Sir Anerood Jugnauth 2003 - 2005 Paul Bérenger

28 Jugnauth first came to power in 1982 as a member of the MMM, but after internal disagreements with the party leader, Paul Bérenger, he split from the party and formed the MSM

2005 - 2014 Dr. Navin Ramgoolam 2014 - 2017 Sir Anerood Jugnauth 2017 - Incumbent Pravind Jugnauth

● Labour Party ● Militant Socialist Movement ● Mouvement Militant Mauricien

Fiji Prime Ministers 1970 - 1987 Ratu Sir Kamisese Mara 1987 - 198729 Timoci Bavadra 1987 - 1992 Ratu Sir Kamisese Mara 1992 - 1999 Major General Sitiveni Rabuka 1999 - 2000 Mahendra Chaudhry

2000 - 200630 Laisenia Qarase 2007 - 2014 (Military Government) Commodore Frank Bainimarama 2014 - Incumbent Commodore Frank Bainimarama

● Alliance Party ● Labour Party ● Soqosoqo ni Vakavulewa ni Taukei (Fijian Political Party) ● Soqosoqo Duavata ni Lewenivanua (United Fiji Party) ● FijiFirst

29 The Bavadra Government only lasted for 31 days before being overthrown in the first coup of 1987 30 For the first year of his term Qarase was not a member of any party, but ahead of the 2001 general election he formed the United Fiji Party