IDENTITY POLITICS AND COMPLEXITIES OF CONFLICT RESOLUTION IN

Lupa Ramadhani

IBIS Discussion Paper No. 2

IDENTITY POLITICS AND COMPLEXITIES OF CONFLICT RESOLUTION IN ZANZIBAR

Lupa Ramadhani

No. 2 in the Discussion Series: Patterns of Conflict Resolution

Institute for British-Irish Studies University College Dublin

IBIS Discussion Paper No. 2

ACKNOWLEDGEMENTS

The author acknowledges funding from the Irish Research Council for the Humanities and Social Sciences and from the Conflict Resolution Unit of the Irish Department of Foreign Affairs.

ABSTRACT

IDENTITY POLITICS AND COMPLEXITIES OF CONFLICT RESOLUTION IN ZANZIBAR

This paper analyses the pattern of conflict resolution in Zanzibar. Since the introduction of multiparty politics in 1992 this semi-autonomous territory within the State of Tanzaniahas remained on the brink of conflict. The paper argues that the conflict in Zanzibar should not be seen as merely a political stand-off with post- election rioting. In fact it has most of the characteristics of a deep-seated and protracted conflict. The political divisions are superimposed on deeper racial/ethnic divisions embedded in territorially-defined horizontal inequalities. These in turn have resonances to very brutal periods in Zanzibar and African history (particularly the slave trade).The paper examines strategies employed in resolution of the conflict to find explanations for the failure of the first and second Muafaka (Agreement). The paper presupposes that the 2010 Reconciliation (Maridhiano) offers actorsbetter chances for the claim to identity than the previous attempts.

BIOGRAPHICAL INFORMATION

Lupa Ramadhani is an assistant lecturer at the University of Dar es Salaam, . He obtained a Master of Arts (2002) in International Relations and is currently a PhD candidate in the UCD Global Human Development ―Sandwich‖ PhD programme. Publications include ―Peace and Conflict in the Great Lakes Region Since 2004‖, Conflict Trends (ACCORD), Vol. 2, 2005, co-authored with Prof. Mwesiga Baregu.

IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

IDENTITY POLITICS AND COMPLEXITIES OF CONFLICT RESOLUTION IN ZANZIBAR

INTRODUCTION Zanzibar is an autonomous region within the state of Tanzania with its own executive, legislature and judiciary. Since 1992, immediately after multi-party politics was introduced in Tanzania, the same party system that prevented conflict on the Tanzanian mainland led in Zanzibar to deep political division on a territorial basis and a cycle of recurrent violence following elections. On deeper analysis, what appears as merely a political conflict with post-election rioting has in fact most of the characteristics of a deep-seated and protracted conflict. The political divisions are superimposed on deeper racial/ethnic divisions embedded in territorially-defined horizontal inequalities (political and economic). These in turn have resonances to very brutal periods in Zanzibar and African history (particularly the slave trade).

The conflict in Zanzibar is of general interest because it exhibits many of the features that make up protractedviolent ethnic conflict: a brutal past history; ethno- racial divisions aligned with a strong territorial basis exacerbated by extreme economic and political inequalities and a politicization of these divisions. However, in contrast to similar conflicts elsewhere in Africa, this conflict appears now to have been settled after only relatively minor violence. It thus poses several puzzles: Was this ever a serious conflict? If so, how was further development of conflict prevented? What role did the Union state play in both conflict and settlement? Did belonging to a stable federal state—Tanzania—crucially help in the settlement process or, conversely, did federalization serve to prolong conflict?

This paper proceeds by sketching the different levels of conflict and settlement processes. It begins by outlining the geo-historical roots of the conflict and the ways in which these developed in the immediate post-colonial period, and subsequently after the revolution and Union. It demonstrates the way in which older tensions and antagonisms were remoulded in the new situation, with a changing politico- economy and new horizontal inequalities which disadvantaged the previously advantaged groups. With multi-party democracy this became politicized not on an explicitly ethnic or racial basis but on a territorial basis. However, within Zanzibar it is clear to all that this territorial basis carried with it strong ethnic and racial resonances and this isshown , for example, in the interrelationships between adherents of the different parties in everyday life (not going to each other‘s weddings and funerals, boycotting businesses run by others and expressing a desire to secede). Of course, this cultural-historical basis is hidden by the Tanzanian party system which is specifically designed to preclude ethno-territorial divisions. However, these divisions are evident in everyday as well as political interactions in Zanzibar. Of course this is not ―ethnic‖ division in any simple sense: several different self-identified groups (Arabs, Shirazi, and Africans) co-exist in complex alliances, now predominantly defined in territorial terms, all speak the same language and almost all follow the same religion. But it is in this manner that ―ethnic division‖ tends to be defined politically in other parts of Africa (see for

- 1 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

example Langer‘s paper in this volume).This paper will trace how pre-Revolution identities were given a renewed salience and formed into novel alliances during the post-Revolution and post-Union period through a very modern politics of distribution and within what appears on the surface as a non-ethnic party system. This is not a repetition of the previous system; rather the old distinctions give added resonance to the new political arrangement.

GEO-HISTORICAL ROOTS OF THE CONFLICT The conflict in Zanzibar has its origin in the long and troubled history of the Isles. Made up of Unguja and Pemba lying 40 kilometers off the coast of Tanzania with a population of a million people (of which 60% reside in Unguja: URT, 2003), Zanzibar has passed through at least three significant phases that together define, and had profound impact on, current socio-political relations and conflict. The first phase starting roughly in the 12th Century coincided with early immigration to Zanzibar by the peoples of the Persian Gulf—mainly Arabs—who established trade links with east African coastal towns, erected garrisons to defend themselves and introduced Islam to the natives (Newbury, 1983). Early immigrants freely intermarried the indigenouspopulation, thus giving rise to a distinct coastal community.

The second phase coincided with the arrival of Portuguese in the 16th Century. The Portuguese established friendly relations with local rulers, set up trading stations and established the Christian mission, secured by Fort Jesus in Mombasa. Portugal became the first European power to gain political control of Zanzibar and subsequently retained control for almost 200 years (Newbury, 1983). Despite this lengthy period of colonization Christianity did not gain firm roots in Zanzibar or other coastal towns.

The third phase may be said to have coincided with Arab control of Zanzibar. Beginning in 1698 after they overran Fort Jesus, Arab forces ejected the Portuguese from Zanzibar and from all other coastal regions North of Mozambique (Romero, 1986). The first and second phases were politically less consequential although highly socially significant as they gave rise to a distinct identity. The second two phases were characterized by immigration of peoples from across the continent. By and large socio-political relations were to radically change in the third phase of Zanzibar history. The Arab connection with Zanzibar grew to such an extent that by 1840 Seyyid Said bin Sultan al-Busaid moved his capital from Muscat to in Unguja (Groot, 1953). Zanzibar Sultans controlled a substantial portion of the East African coast known as Zanj, and trading routes extending much further across the continent, as far as the present day Democratic Republic of the Congo. The Arabs also opened huge plantations that depended largely on slave labour and later on squatters. To pave the way for plantation most of fertile and arable land was taken from Africans; thus sowing the seeds of future socio- economic tensions. Unlike the first and second phases, the phase of Arab rule led to massive immigration of Africans from the Mainland especially along the slave routes. Africans therefore would immigrate to Zanzibar to take up the low social status positions—mainly as slaves—to work in Zanzibar plantations.

- 2 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

Arab rule established extensive links with European powers—many of them already colonial powers on the Mainland—and this was reflected by the number of consulates established. In the 1890s Anglo-German rivalries already were too high to threaten the survival of the Arab rulers in Zanzibar. Indeed ―all sultan successors owed their control over the throne to implicit or overt European intervention‖ (Newbury, 1983: 256).

British rule was established in Zanzibar in the late 1890s mainly to oversee the abolition of the slave trade and to check the spread of German influence over the East African coast. Indeed in the 1890s to counter the increasing encroachment of German power, the sultan requested the establishment of a British protectorate over Zanzibar (Newbury, 1983). With the advent of British rule a system of dual colonialism emerged in Zanzibar since the British established ―indirect rule‖ which assured the Sultan and Arab aristocracies a big share in the running of the new British bureaucracy.

DEMO-POLITICAL DYNAMICS IN ZANZIBAR The presence of diverse groups superimposed on exploitative economic relations coupled with an exclusive colonial political system became a source of political turmoil. The much cited population census of 1948 was the last official and systematic enumeration to carry a complete study of group identities. It indicated four principal groups including Arabs, Asians, Africans (from the mainland and indigenous) and others (including Comorians, Goans, and Europeans) (Newbury, 1983; Lofchie, 1963). The significance of the figures in Table 1 relates to the proportion of the Arab population—standing at 17%—which has remained a significant factor in socio-political relations to date. Apart from South Africa— comprising 20% of non-African population—Zanzibar had the second largest non- African majority in Sub Saharan Africa (Lofchie, 1963).

Table 1: Ethnic/Racial composition of Zanzibar population, 1948 Census UNGUJA PEMBA TOTAL RACE NUMBER % NUMBER % NUMBER % African 118,652 79.3 81,208 70.9 199,860 75.7 Arab 13,977 9.3 30,583 26.7 44,560 16.9 Indian 13,107 8.8 2,104 1.8 15,211 5.8 Comorian 2,764 1.8 503 0.4 3,267 1.1 Goan 598 0.4 83 0 681 0.3 European 256 0.2 40 0 296 0.1 Other 221 0.2 66 0 287 0.1 Total 149,575 100 114,587 100 264,162 100 Source: Lofchie (1963)

The 1948 census is a reflection of centuries of migration to Zanzibar of different ethnic groupings, which in turn were placed into a hierarchical socio-political system. Africans from the mainland arrived mainly as slaves or porters; Arabs were slave traders and aristocrats; Europeans mainly as expatriates and later on colonialists; and Indians as traders concentrated in the urban centres in Unguja (Sheriff, 2001). As will be highlighted later in this paper, geographical factors contributed to shape identity and politics in Zanzibar. It is demonstrated, for

- 3 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

example, in Table 1 that there were twice as many Arabs in Pemba as in Unguja and the number of Africans in Unguja was one and a half times the number of Africans in Pemba.

The figures and categorization in Table 1 have not remained constant, as identity was instrumentalized relative to benefits of belonging to certain categories. These categorieswere neither rigid nor impervious. Sheriff (2001: 308), for instance, has shown that between 1924 and 1948 Swahili who had numbered 34,000 had virtually disappeared by 1931 ―as the label came to be seen as a pejorative term referring to people of slave origin…On the other hand, the Shirazi, as the indigenous people of Zanzibar preferred to call themselves, grew from 26,000 in 1924 to 41,000 in 1934‖. The Shirazi, claiming a mixed blood from the early 12th century Shirazi from the Far East, is largely an identity invention arising out of desperation for the original inhabitants of Unguja and Pemba to differentiate themselves from recent African groups of mainland origin brought in mainly as slave labourers (Newbury, 1983). Clearly, as studies have confirmed, the slave trade and slavery left indelible marks in Zanzibar and are repugnant legacies. In a mid-1980s study of the legacy of slavery, it was found that today very few people associate themselves with slavery despite the fact that at the height of the ―trade‖ the slave population outnumbered all free people combined (Romero, 1986; Killian, 2008). As was evident in Zanzibar, the slave trade disintegrated the core African group. Africans—the largest group in the isles—invented other identities such as ―Shirazi‖ in order to distance themselves from the disgraces of slavery. As Tambila has observed, slavery formed part of the relations of production and resulting class conflict and class relations, though changing over time,italso informed ideological positions deriving from one‘s social class (Tambila, 2001). In the ―sons of soil‖ understanding the Shirazi gave rise to three distinct groups in Zanzibar: waTumbatu, waHadimu, and Pemba Shirazi. It is the way each of these groups allied with the minority but dominant Arab power holders and/or the African majority powerless in the struggle to alter power relations that determined political outcomes in the run up to independence in 1963. The ZNP/ZPPP coalition government came about after a political alliance of the Arabs and Shirazi, thus elevating the discrimination of Africans.

The post-1964 Revolution period, however, ensured a changing political landscape commensurate with noticeable identity shifts deriving from the Africans acquisition of power. Killian (2008) has shown that in the early 1970s the Revolutionary Government of Zanzibar issued a circular and mechanisms to enable people to change their identity. Many who had previously identified themselves as Shirazi now identified themselves as Africans. ―While in 1948 about 56% of Zanzibaris identified themselves as Shirazi, only 42% said so in early 1960s and this number had declined to 20% by the early 1980s‖ (Killian, 2008:106). Clearly in the post Revolution period African identity became safe and advantageous for one to identify with. But interestingly, as freedom of expression increased with the advent of liberal democracy, expression of Arab and Shirazi identity started to re-emerge. For example in a 1999 survey, Killian found out that 51.3% of respondents identified themselves as ―Africans‖, 27.5 as ―Shirazi‖ and 12.7% as ―Arabs‖.

- 4 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

It is important to point out that the Zanzibar-Pemba regional divide coincided with colonial penetration and the class relations that arose from it. By the 1830s the economy of Zanzibar was firmly in the hands of Arabs who had established large coconut and clove plantations, especially in Unguja (Sheriff, 2001; Killian, 2008). Pemba started to attract investments only when world market prices plummeted in the 1870s and extreme weather wiped out many clove plantations in Unguja (Sheriff, 2001). These events are significant in a number of ways. Firstly, the number of African immigrants from the mainland was much larger inUngujain comparison to Pemba. Secondly, the scale of land alienation was noticeably higher in Unguja as a result of increased immigration.Thirdly, following this, the level of Arab-African antagonism was of a more intense nature in Unguja compared with Pemba. According to Sheriff (2001) Pemba attracted relatively fewmainland immigrants and Arab-African/Shirazi socio-economic relations were in turn more harmonious. Since the shifting economic attention to Pemba coincided with (i) the abolition of slave trade and (ii) a global decline in price of cloves, by the time Arabs started clove farming in Pemba in the 1870s they had to rely mainly on free labour. Moreover, since they were relatively impoverished they could not establish large plantations, relying therefore on small plots maintained on a feudal basis. According to Sheriff (2001) Arabs were forced to enter into agreement with Africans whereby Africans would help in the clearing of the fields in the understanding that they would gain half the land after a predetermined period. Politically, therefore, Africans in Unguja

were struggling against Arab hegemony and wanted to restore their expropriated land and local autonomy while in Pemba the Shirazi regarded the mainlanders as enemies and foreigners accusing them of taking their jobs and promoting Christianity in a predominantly Muslim state (Killian, 2008: 105).

Groups were thusstrongly heterogeneous and belonging to one group did not stop one from re-categorizing depending on economic circumstances. Of course there were poor Arabs much the same as there were rich Africans although this rarely affected the political alliances that were formed.

It seems reasonable to suggest that associational life is very important in holding society together. In a study of ethnic conflict and civil society in India, Varshney (2001) divided community life into associational forms and everyday engagement and observed that associational forms of engagement have a stronger impact on political relations than everyday engagement and are more prone to political manipulation than the latter. Political parties in Zanzibar, as in many parts of Africa, seem to follow this pattern. In the 1920s, largely as a response to the impact of the Great Depression on local populations, four prominent associations emerged: the Arab Association, the African Association, the Shirazi Association and Indian National Association (Sheriff, 2001). The British colonial policy encouraged intragroup solidarity and mobilization. Arabs were favored as a ruling group and have since been the main driving force in the colonial bureaucracy and policy organs. As far back as 1914 only the British, Arabs, and Indians had representation in the Protectorate Council. Even after the establishment of the Legislative Council in 1926 representation remained organized in this way (Killian, 2008). Africans of

- 5 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

Shirazi identity were appointed for the first time into the LEGCO in 1946 signaling that the British defined the Shirazi as the indigenous Zanzibaris rather than African of Mainland origin. This led to further polarization of the African group as the Shirazi argued that generational longevity in the Islands had to be used as the primary criterion for representation (Killian, 2008).

THE CONJUNCTURE: THE ETHNO-NATIONALIST CLEAVAGES IN THE INDEPENDENCE STRUGGLE As in many other parts of Africa, anti-colonial struggle in Zanzibar gained momentum after the Second World War and especially in the 1950s. As Mukangara (2000) observes, by the mid-twentieth century Zanzibari communities were already racially and ethnically divided. Racial relations characterized political affiliation and became the driving force for formation of political parties. In a climate of increasing tensions the African majority were bitter about sociopolitical and economic deprivation at the hands of an alien minority (Campbell, 1962; Mukangara, 2000).

Political parties appearing in the 1950s became polarized along racial lines, reinforcing ethnic and regional divisions. All political parties could be traced back to previous racially based groups formed mainly in the 1920s to cater for specific interests of the groups. Africans and Shirazi (of Unguja) formed the Afro Shirazi Party (ASP) in 1957; Arabs formed the Zanzibar Nationalist Party (ZNP) in 1955 and the Shirazi of Pemba splintered from the ASP to form Zanzibar and Pemba People‘s Party (ZPPP) in 1959; henceforth voting in coalition with the ZNP (Mukangara, 2000). It is interesting that even the Shirazi and Africans of Unguja only came together to form the ASP after a visit by Nyerere, then leader of TANU in Tanganyika, to Zanzibar in 1957 to encourage them to unite (Campbell, 1962). The pattern of support of these parties clearly followed the racial and ethnic location of the various groups. Africans of mainland origin, for instance, were mainly found in Unguja where the ASP had a stronghold. The ZNP drew its support from the areas where the Arab elite lived in Ungujaand the heavily Pemba based ZPPP drew support from Pemba. Although the Indian National Association was not transformed into a political party, it is clear that Indians were more likely to vote in alliance with Arabs than Africans since they exhibited a closer relation to power elites than to Africans. This combination of factors led to a party system that was almost balanced down the middle.

Issues of ethnicity and racial division dominated political campaigns with parties desperately struggling to win popular support. While ASP clearly stated that it stood for an African majority commensurate with an African government for Zanzibar, ZNP on the other hand advocated mechanisms to safeguard the monarchy and build an Islamic multiracial state. They accused ASP to be ―alien‖, a thinly veiled referral to the party‘s supportfrom recent immigrants from the Mainland. A religious chord was struck in the claim that if it was to be elected ASP would Christianize Zanzibar (Killian, 2008).

Four elections were organized by the British government from 1957 to 1963 and a clear pattern may be deduced. Firstly, it was evident that none of the political

- 6 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

parties could command an absolute majority to enable them to formulate a government. Therefore it was important for parties to work together in coalition government. Secondly, election results were highly contested mainly due to questions of legitimacy over the electoral management body. This is seen in several instances of shifts of constituency boundaries with a covert gerrymandering motive. The number of seats was increased from a mere 6 in 1957 to 31 in the July 1963 election. Thirdly, and unfortunately, Africans and Arabs failed to surmount their differences and work together in a coalition government, the only alternative that could logically ease the tensions. The 1957 election—the first to be held in Zanzibar—was the only one to be clearly won by the ASP who gained 5 of the 6 seats in the Legislative Assembly, enabling Karume to form an internal self- government (Rwambali, 2001). Even with this clear victory political discontent was high as many eligibility conditions were stringent; hence only 40% of the expected 100,000 people were eligible to vote. Following accusations of double voting riots broke out in the Darajani area, the heart of Unguja town and strong ZNP base. Voters who were suspected of twice casting a ballot or of voting from other districts were attacked with stones and knives leaving 62 people dead and scores more injured. A state of emergency was declared with British troops from Kenya and the Middle East called in to restore order (Campbell, 1962).

Table 2: January 1961 Legislative Council Election PARTY UNGUJA PEMBA NATIONAL TOTAL VOTES % SEATS VOTES % SEATS VOTES % SEATS ASP 27,495 61.11 8 9,203 23.03 2 36,698 43.19 10 ZNP 15,555 34.57 5 17,169 42.96 4 32,724 38.52 9 ZPPP 1,946 4.32 13,595 34.02 3 15,541 18.29 3 TOTAL VALID 84,963 22 SOURCE: AFRICAN ELECTION DATABASE

The next elections in January 1961 were inconclusive, as Table 2 indicates, since neither ASP nor ZNP were initially able to form government. Only later was a compromise reached to form a coalition government. Of interest to the January 1961 elections is the relative regional and racial division between ASP on the one hand and ZNP and ZPPP on the other. While ASP enjoyed 61% of the popular vote in Unguja, it managed only 23% and 2 seats in Pemba. Conversely, ZNP emerged with 43% of the Pemba vote and only 35% of Unguja vote.

Rivalries, disagreements and violence followed the June 1961 Legislative Council Elections. At the root of the contestations was a constitutional crisis accompanied by an emergent nationalism that demanded for radical changes to incorporate Africans in government. The depth of the crisis was evident following six days of ―bloody rioting‖ (Lofchie, 1963: 186). Indeed as Lofchie correctly argues, all political parties in Zanzibar professed a common ideology of nationalism, but had irreconcilable differences between them that led to a legacy of bitterness and intransigence that in turn acted as a centrifugal force within Zanzibar society (Lofchie, 1963). However, at the centre of Zanzibari nationalism lay a deep contestation as to the identity of the Zanzibar state itself as witnessed by the

- 7 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

campaign platforms of political parties: Would it be an Arab state? Would it be a Shirazi or African State? Political parties saw different routes to a Zanzibar state thereby exacerbating the threat of conflict.

1963 saw the last pre-independence elections. The universal electoral suffrage almost doubled and the electorate and the number of seats was raised to 31 (Sheriff, 2001). The ZNP and ZPPP formed a coalition and with 18 seats in the National Assembly, assembled the first post-independence government on 10 December 1963. Interestingly, the Sultan assumed the role as head of state. ASP immediately cried foul. It had garnered 1,092 more popular votes that the ZNP and the ZPPP combined and 54% of popular vote. ASP complained that this was a plot by the colonial government through the gerrymandering of electoral districts to stop it from translating its majority into seats (Mukangara, 2000).

Table 3: July 1963 National Assembly Election PARTY/COALITION UNGUJA PEMBA NATIONAL TOTAL VOTES % SEATS VOTES % SEATS VOTES % SEATS ASP 53,232 63.09 11 33,853 49.39 2 87,085 54.21 13 ZNP/ ZPPP 31,144 36.91 6 42,415 55.61 12 73,559 45.79 18 ZNP 26,572 31.49 6 21,378 28.03 6 47,950 29.85 12 ZPPP 4,572 5.42 21,037 27.58 6 25,609 15.94 6 TOTAL VALID 160,644 SOURCE: AFRICAN ELECTION DATABASE

As Table 3 indicates, the election of 1963 was yet another indication of the polarization of votes among political parties and their bases of power. It reconfirmed the ASP‘s strength in Unguja (63%) relative to Pemba (44%). It was an improvement in that ASP was able to close the gap in Pemba compared to previous elections although like 1961 it ended with only two seats in Pemba. ZNP appeared to have an even distribution of votes in Pemba and Unguja with 31% and 30% respectively. The distortion only comes when ZNP and ZNPP votes are collapsed in Pemba. ASP was two times more influential in Unguja than the ZNP. But again ASP had a respectable lead even in the ZNP‘s stronghold.According to ASP the procurement of independence in 1963 therefore was translated as Arab Independence (Killian, 2008).

THE 1964 REVOLUTION AND THE UNION: TURNING BACK THE CLOCKS With the numbers on their side and the feeling that the 10 December independence was ―Arab Independence‖, ASP was left with no alternative except to wage a revolution. The Revolution, said to have been carried by not more than 600 insurgents, targeted the ruling aristocracy including Arabs and Indians (Daly, 2009). Barely four weeks from assuming power, on 12 January 1964, the Sultan had been overthrown by the African forces. There was little resistance as the revolutionaries‘ overpowered the police with traditional weapons, capturing arms and reinforcing their ranks. The sultan and members of his government fled the country. SheikhAbeidAmani Karume was named President of the newly created People‘s

- 8 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

Republic of Zanzibar and Pemba. Several thousand Arabs (5,000-12,000) and Indians were killed, thousands more detained or expelled, their property either confiscated or destroyed. Large landholdings were nationalized and distributed to the landless squatters in three-acre plots. Trade was made a state monopoly (Sheriff, 2001). In addition 611 homeswere confiscated, many in Stone Town which was a residential area for Arabs and Indians (Killian, 2008).

After the Revolution the ASP government immediately nullified the constitution and banned all political parties except the ASP. Civil society activities were banned as well. The 10 December is never celebrated as Independence Day; rather it is 12 January, the Revolution Day that is celebrated in colorful annual ceremonies. By any means, as already explained, the 10 December remains independence day for Arabs. The Revolution effectively marked the end of liberal democracy and paved the way for rule by decrees of the Revolutionary council.

The Union with Tanganyika barely three months after the revolution on 26 April 1964 has remained a subject of much speculation. It is believed that Karume never felt secure and feared retaliation from the overthrown forces that could easily regroup. Pemba for instance remained lukewarm to the Revolution. Being geographically isolated from Unguja with the largest Arab population in the Islands, it posed a genuine political threat. There were also within the ASP intragroup squabbles and radical elements that Karume needed to neutralize. This could plausibly explain the mysterious disappearance of John Okello and other left wing elements. The Union with the Mainland, which had strong historical and blood links with the Africans in Unguja, would provide much needed political relief. Many of the radical elements within the ASP were therefore either transferred to Tanganyika where they served in the Union government or were suppressed. It is also speculated that the Western bloc led by the US was worried about the connection between some elements within the new regime in Zanzibar and the Communist bloc; rumours of militant factions having received training in China, USSR and Cuba. The earliest countries to recognize Zanzibar were Communist governments; Zanzibar simultaneously offering recognition of East Germany then not recognized by countries of the Western Bloc.

The Union complicated the Unguja-Pemba relations in several ways. First it affirmed the long held hypothesis that ASP was formed by ―aliens‖ and this made Pemba even more sceptical. The question of the legitimacy of the Union has been a recurrent theme ever since and has remained a delicate subject. Some of the contentions surrounding the Union include inadequate consultation prior to the Union, structural defects, difficulties of sharing costs and benefits and an increased demand for autonomy (Killian, 2008). If the Union was entered into without sufficient knowledge of the ruling elites in both sides, it has been perpetuated through even more repressions. Public debates over the form or legality of the Union were regarded as treasonous and some politicians got victimized for questioning the Union. , assuming the presidency in the aftermath ofKarume‘s death in 1972, was forced to resign in 1984 ―because of his intention to call for a Special Constitutional Court to determine the fate of Zanzibar‖ (Killian, 2008:112). Four years later SeifShariffHamad, then Chief Minister, was also

- 9 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

expelled from the party and forced to resign partly because of his pursuit of autonomy of Zanzibar. Killian (2008) sees this as complicating the ethno-political relations in Zanzibar. Since the establishment of the Union, another struggle has emerged; that of restoring the identity of the Zanzibar state. Very interestingly, demands for Zanzibar autonomy bring together Zanzibaris regardless of regional or ideological affiliation.1 More curiously, however, some people see the move to create more autonomy for Zanzibar as paving the way for Pemba to declare its independence. In fact while CCM position is a two government structure, CUF advocates for a three tier loose federal structure which makes it easy for constituencies to break away.

THE POST UNION PARTY SYSTEM The Union as elaborated above brought together two sovereign states— Tanganyika and Zanzibar. At the time of unification Tanganyika was already a de facto one party state. In 1961 the British handed over power to the Tanganyika African National Union (TANU) having won all the seats in the Legislative Council. From 1964 until 1977 when the two parties merged, TANU remained the sole political party in the Tanzanian Mainland as it came to be known while ASP remained the party for Zanzibar. In February 1977 TANU and ASP merged to form (CCM) unofficially translated as the Revolutionary Party. For all practical purposes, and in relation to the critical aspect of identity, CCM claim to be the heir to ASP and custodian of the Revolution and the Union. In terms of current labeling CCM is still dubbed ―a party of Mainlanders‖ by their main opponent, the Civic United Front (Mbunda, 2009). CUF alongside several other political parties, on the other hand, was formed in 1992 after the re-introduction of multi-party politics in Tanzania. Currently (November 2010) there are 18 fully registered political parties in Tanzania.

Although according to the 1992 Political Parties Act in Tanzania, political parties are supposed to espouse national unity and refrain from sectarian politics in order to qualify for registration; that does not stop constituencies from supporting one political party over another depending on several factors. But as we shall see later in this paper, what is striking in the Zanzibar political party system is the resurgence of the pre Revolution patterns of political support. For instance, CUF performs well in Pemba and generally in areas where ZNP and ZPPP were successful while CCM strongholdscoincide with ASP‘s.

Furthermore, CCM and CUF in turn use the past to boost popularity and dent the rival parties. CUF has at least three identities and all of them problematic. Firstly, it is associated with all those who were either oppressed or dissatisfied with the way the Revolutionary Regime worked and wanted radical reforms. Secondly, it is

1The debate on the statehood of Zanzibar recently resurfaced when the Prime Minister of Tanzania in a question and answer session in parliament responded that Zanzibar was not a state. This statement was angrily received in Zanzibar where the HoR took considerable time to discuss it. The Union president eventually cooled off the matter by stating that Zanzibar was a ―state‖ in so far as all non-Union issues were concerned, and not a state in all other Union affairs including currency and Citizenship. In 1992 for example, Zanzibar unilaterally joined the Organization of Islamic Conference (OIC) but was forced to withdraw in 1993 since it was argued that Zanzibar government did not have mandate to join international organizations.

- 10 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

labeled, especially by CCM, as representing ZPP and ZNPP; the parties that were overthrown by the current regime. Thirdly, and related to this, CUF is seen to be the champion of ―Pemba nationalism‖ (Mbunda, 2009), an identity which they carefully try to avoid. Pemba nationalism has ebbed and flowed but was strongly expressed in April 2008 when Pemba elders and CUF leaders met with the Ambassador of United States of America to Tanzania, Mark Green, and requested the USA to help Zanzibar gain independence from the United Republic of Tanzania ―as the USA did for citizens of Kosovo, who successfully seceded from Serbia‖ (Mbunda, 2009: 5). Although Police arrested and kept in custody 12 elders for possible charges of ―treason‖, no charges were formerly registered and the elders were subsequently released. The significance of these pronouncements lay not in the legal weight they carried but in conveying the message of an area that felt neglected by the regime.

THE SKETCH OF THE CONFLICT The Zanzibar conflict reflects what Todd (2005: 93) describes as an ―extreme case where the state is simply the instrument of the dominant community, its interests constituted by community interests—an ‗ethnic state‘‖. Indeed in Zanzibar the dominant group has been using state resources to secure and reproduce its own dominance and to exclude the dominated group. The Revolution was poised to affect Pemba negatively given the kinds of political and economic policies that the government adopted. As Sheriff argued ―by declaring Zanzibar a one-party state and banning the overthrown political parties, the revolution essentially disenfranchised nearly half the population‖ (Sheriff, 2001: 315). Some examples follow to help explain why Pemba started to feel a sense of alienation which slowly pushed it to seek a separate collective identity. An opportunity to wage a struggle for its reclamation emerged following the re-introduction of multi-party politics.

Primarily in Pemba, political and economic inequalities go together and actually reinforce each other. People of Pemba enjoy fewer benefits from the state than fellow Ungujans. All of the six presidents and many of the chief ministers in Zanzibar for over four decades came from Unguja. In a country that is regionally divided this created a powerful sense of alienation. The fact is that that only after the Reconciliationof 2010 (Maridhiano)—to be detailed later—did the first president from Pemba come to be elected following the 31 October 2010 election in Zanzibar. , the only Chief Minister from Pemba, who was ascending the ladder into the presidency, was suddenly dropped from the cabinet in 1988, expelled from CCM and detained for 30 months based upon accusations of tampering with the Union (Killian, 2008). Incidentally, Hamad is now first vice president in the Government of National Unity that was formed after Maridhiano. Since the cabinet is drawn from the winning party, Pemba ended up with very few cabinet posts. The following extract from an interview with Nassor, a resident in Pemba, captures the feeling of the common people in Pemba….―the infrastructure here is so poor, unemployment rate is higher than in Unguja…and despite its size and population, the entire island has only two ministers in the government..this is unfair‖ (Daily News, 26th May 2010). Similarly, the feeling is that Pembans are excluded in other government posts down the bureaucracy. CUF maintains that in

- 11 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

the previous government (in power from 2000 to 2010) Pemba had less than 20 percent of the bureaucratic posts in the government (Mbunda, 2009).

Pemba is also visibly marginalized economically. The general state of infrastructure (roads, electricity supply, tourist hotels, fast ferries, number of air flights) is poorly developed in Pemba compared to Unguja.Tourists fly directly via Zanzibar international airport to Unguja and stay in numerous luxurious beach hotels. Tourists rarely visit Pemba because there is no physical infrastructure, and many who visit ―have been shocked by obvious signs of malnutrition among children‖ (Sheriff, 2001: 315).

Relatedly, the Household Budget Survey (HBS) of 2004 revealed that of the 94 industrial establishments in Zanzibar, only 15 were located in Pemba. Many of them had less impact economically as they were small scale with 9 employing less than 10 people and only 3 employed between 50-99 people. The rest of the industrial establishments were located in Unguja with 3 alone employing up to 500 people. The HBS established consumption per member of household to find out poverty levels in Zanzibar. The results indicated that 13% of people in Zanzibar lived below the poverty line, with a further 49% unable to meet their basic needs. The survey also uncovered the rural/urban dichotomy in the incidence of poverty with rural areas having a much higherrate than urban areas. Since Pemba is still heavily rural in comparison to Unguja, this implies that Pemba is more affected by poverty than Unguja. The HBS found that while the total percentage of the population ofUngujaconsidered poor amounted to 10%, in Pemba the figure approached to upwards of 20%.

As a major world producer of cloves (70% of global output), Pemba has been less satisfied with the way the Zanzibar government handles the economy and complained about the government‘s exploitative monopoly over the clove market, at times offering 50% less than world market prices (Pottie, 2002), or in the worst cases as in 1978 only 7% (Sheriff, 2001). The Zanzibar government‘s Special Forces KMKM (coastguards) ensure that clovesare not smuggled to nearby Kenya where prices are far better (Sheriff, 2001). Moreover, smugglers are treated witha heavy hand. While Pemba claims to be generating a large share of the national income, the income seems to be benefiting Unguja.

THE RISE OF VIOLENCE AND REPRISALS When in 1995 violence erupted in Zanzibar following disputed election results, people outside Tanzania had little notion of what was going on. Africa was grieving from an unprecedented stream of violent conflicts. Genocide in Rwanda in 1994 had left a million people dead and thousands of refugees. Its effects spread violence throughout the Great Lakes region mainly in Burundi and Democratic Republic of Congo (DRC). If weighed against the damage and destruction of the Great Lakes conflicts, Zanzibar seemed insignificant in scale. This explains its obscurity and the general indifference of the international community to address it. The international media, known at times for picking only trouble spots in Africa, was simply busy elsewhere. But the conflict in Zanzibar was real and has tended to

- 12 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

grow and become more complex and hence more difficult to resolve. What might differentiate Zanzibar from other countries that entered into large scale conflicts is the tactics utilized. Escalation did not reach the level where military struggle became an option. Moreover, the Union government acted as a stabilizing force much the same as the United Kingdom would provide stability in the Northern Ireland conflict.

Violence in Zanzibar has tended to erupt during election time. Four general elections have been conducted in Tanzania since independence and the trend is surprisingly similar to that of the 1957-1963 elections in Zanzibar, only that the power balance has shifted. The bottom line has been the apparent fear of the Revolutionary government to lose power as ―this government came about through struggle and blood and will not be let away through a piece of paper‖.2 CUF on the other hand is aware of such sentiments and feels that its rightful democratic victory was deliberately denied. Indeed elections have tended to act as sparks of violence in Zanzibar. As correctly argued by Cameron, CCM leaders and many people in Unguja think that CUF will return the Omani Sultanate, associating CUF with the pre-Revolutionary ZNP which was overthrown by ASP. A speech by the first lady at the meeting in Pemba epitomized the CCM‘s campaign:

opposition leaders dreaming of a return the pre-Revolutionary regime ... before the 1964 Revolution the people of the Isles were slaves in their own country... I call upon you to be more analytical before you support parties with hidden agendas aiming at humiliating your dignity as independent people (Cameron, 2002: 316).

As a result, CCM employed similar oppressive strategies to that of its predecessor—the pre-Revolution government—to stop CUF gaining access to power. Table 4 and 5 below indicate the closeness of political competition in Zanzibar and reveal a yawning gap in the current winner-take-all electoral system. In the presidential elections, for instance, the difference between the winner and loser is just a fraction of a percentage.3 The 1995 elections were held amidst claims for more meaningful reforms including a new constitution and reformulation of the electoral commission. It is argued that CCM played a precautionary game allowing only symbolic reforms while the previous superstructure and machinery of a one party state remained intact. In such an environment CCM won narrowly the presidential seat in an election that was full of fraudulent practices. CCM won 50.2% of popular votes against CUF 49.7% and gained 26 of the 50 seats in the House of Representatives. CUF refused to recognize the government and boycotted all sessions of the House of Representatives. They maintained that the elections were rigged to rob them of victory. Hamad declared himself the rightful elected president of Zanzibar. The CCM government responded by arresting 18 CUF followers keeping them in custody without in what was coined ―treason trial‖.

Table 4: Elections for Zanzibar House of Representatives (1995-2010)

2This was a message that was circulated during the 1995 election by anonymous people. 3Elections on the Tanzanian Mainland and Zanzibar are usually held on the same day. In Zanzibar, however, in addition to the Union President and Members of the Union Parliament, voters elect the Zanzibar president, members of House of Representatives and local councillors. The HoR has 50 elected members.

- 13 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

UNGUJA PEMBA YEAR PARTY Votes % of votes Seats Votes % of votes Seats CCM 134,399 86.1 26 21,632 13.9 - 1995 CUF 40,212 25.5 3 118,716 74.7 21 CCM 2000 CUF CCM 203,713 89.7 31 23,346 10.3 - 2005 CUF 82,117 41.5 1 115,693 58.5 18 CCM 22 - 2010 CUF 4 18 SOURCE: KILLIAN, 2008; ZEC ELECTION REPORT 2010

In the 2000 elections CCM had a new presidential candidate, Amani Karume, who had to contest with Seif Sharif Hamad who was running for the second time having been contentiously defeated in 1995. Karume gained more popular votes than his predecessor (67% of the popular vote cast) in an election that was again full of controversy and recriminations.

Table 5: Zanzibar Presidential Election Results (1995-2010) ELECTION PARTY TOTAL % %UNGUJA %PEMBA YEAR VOTES 1995 CCM 165,271 50.2 87.4 12.6 CUF 163,706 49.8 39.0 61.0 2000 CCM 248,095 67.0 - - CUF 122,000 33.0 - - 2005 CCM 239,832 53.2 90.6 9.4 CUF 207,773 46.1 42.0 58.0 2010 CCM 179,809 50.1 63.9 18.2 CUF 176,338 49.1 33.4 80.1 SOURCE: KILLIAN, 2008; ZEC ELECTION REPORT 2010

The 2000 elections were described as being more corrupt and more rigged than those in 1995. Local and international election observers reached similar conclusions that elections were neither free nor fair and could not represent the wishes of the Zanzibar people. A Local observer group, TEMCO, simply stated that elections were ―aborted‖. The Zanzibar Electoral Commission (ZEC) grossly mismanaged the vote counting process as vote counting was halted in over 40% of the polling centres; ballot boxes sent to District Commissioners‘ offices without the presence of party agents. This gave CUF more impetus to claim for a re-election and grounds to boycott the new regime. CUF refused to recognize Karume as president of Zanzibar, organized countrywide boycotts and in one incidence ―police and protestors clashed in Stone Town. Police fired shots at protestors and 16 CUF supporters were arrested. The nearby market and some streets were closed‖ (Pottie, 2002: 347).

The general situation remained explosive and delicate. ZEC offices, churches and pubs were targeted for attack normally by home-made bombs. CUF and CCM traded accusations with each side placing blame on f the other for the explosions. While CUF maintained that explosions were a typical CCM ploy in order to gain an excuse to crack downon CUF members, CCM accused CUF of using ―guerrilla‖

- 14 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

tactics to render the country ungovernable. CUF boycotted all sessions of the House of Representatives and the National Assembly. Nationwide demonstrations were organized; albeit most of them banned by the Police. Following police claims of illegal demonstrations in Dar es Salaam CUF Chairman Prof. Ibrahim Lipumba together with several other CUF followers were arrested, roughed up and tortured when in custody. Tensions came to a head on 27 January 2001 when the police crashed with angry demonstrators in Pemba, shot live ammunition randomly at the crowd and in the fracas killed over 27 demonstrators.4 Despite Amnesty International‘s strong communiqué requiring the Tanzanian government to establish and explain what happened, as well as bring to justice any public official responsible for human rights abuses and unlawful action, no visible action was taken (AI, 2001).

Similar eruptions of violence emerged during the 2005 election. At the height of tensions a leader ofthe Valantia5 militia in Pemba was stabbed to death by people who were suspected to have been disgruntled with the role of the armed forces in preventing CUF followers from registering. However, the killers have not been arrested. In April 2005 four people were killed by the police following violence after CUF supporters protested against the refusal to be registered (FEWER, 2005). Violence and hostilities spread to affect other forms of everyday interaction between Zanzibaris. Followers of political parties stopped buying goods from shops of those belonging to rival political parties and refused to attend funerals or other social functions of opposing followers. Furthermore, in Unguja some people of Pemba origin received written threats urging them to leave immediately.

The elections in October 2005 saw CCM emerging victorious against its arch rival CUF amidst the usual cries of foul. As Table 5 above indicates, Karume‘s votes slightly declined from 67% in 2000 to 53% while Hamad managed to increase his votes from 33% to 46%. CCM scored 31 seats while CUF finished with the remaining 19 seats. But what is remarkable is that CUF won all the 18 seats in Pemba and managed one seat in Unguja, while CCM pooled all its seats from Unguja. The historical divide seems to be entrenched. CUF immediately disowned the election results and refused to recognize Karume citing a gross breach of electoral laws, this time pointing to the massive registration of armed forces, double voting and deliberate disenfranchisement of CUF followers. Much like a sadistic movie, the situation in Pemba once again descended into the doldrums.

Having traced the complex development of political party system aligned with ethno-territorial resonances and the violence involved, the paper goes on to explore how conflict in Zanzibar was resolved.

THE SETTLEMENT PROCESS 1995-2010: ACHIEVEMENTS AND DRAWBACKS

4Accurate figures have not been established as the government never established an independent commission to enquiry into the killing. A joint fact finding report by the International Federation for Human Rights and the Legal and Human Rights Centre found out that altogether 65 people were killed in the whole of Zanzibar on 26th and 27th January 2001. 5Valantia is one of the special paramilitary forces of the Revolutionary Government of Zanzibar.

- 15 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

The likelihood for resolving any conflict depends on the extent to which underlying issues are identified and addressed before reaching agreement. In certain cases simple issues are neglected or partially addressed, leaving them to grow into grievances that over time trigger violence. Conflict becomes very difficult to resolve after violence has erupted, as violence takes on a logic of its own. Zanzibar‘s peace process has been a protracted affair; first attempts at settlement were initiated in 1995 before agreement was finally reached in 2010. Several strategies of conflict resolution have been attempted in order to accommodate ethno-political rivalries in Zanzibar. The Commonwealth mediation after the 1995 election is the only direct international involvement and third party participation in the conflict. Apart from the Commonwealth mediation, contending parties have also attempted negotiations. These attempts at forging settlement are usually referred to as ―inter party‖ negotiations and occurred immediately after the elections of 2000 and 2005. The Commonwealth mediation led to the first Agreement (first Muafaka, 1999); the interparty negotiations came with the Second Muafaka in 2001. However, although the two agreements eased the tensions to some extent, they were not able to resolve the conflict.

Settlement, therefore, has to be viewed as a typically cumulative effect of several layers of personality shifts, agreements, and structural transformations. None of these processes could be said to have led to the present outcome alone. This is in marked contrast from cases where electoral violence erupted on a massive scale, prompting international reaction and intervention, andas in the case of Kenya, leading to a quick resolution of the conflict. Zanzibar has remained low key because violence was seen to be less intense and therefore more manageable.

At the individual level it is possible to talk of the executive-level personality shift which, coupled with changing circumstances, seems to have contributed significantly to the attainment of peace. Since 1995 Zanzibar has passed through two different personalities as heads of the Revolutionary Government. Obviously both have faced different sets of circumstances that in turn have led them to view challenges and opportunities posed by the conflict in a contrasting fashion. According to the Zanzibar and Tanzania constitution, the president can only serve for two terms of five years each.

Salmin Amour, coming to power as president first during the single party system, steered the country into plural politics and had to face the first multiparty election in 1995. He is known to have been a hardliner, a man who stood for ―MapinduziDaima‖ (Revolution Forever)—CCM‘s motto in Zanzibar—and as a personality resolute in his refusal to cede too much to the CUF demands. He indeed earned the name of ―Commando‖ not only by the way he dealt with CUF with a heavy hand, but also how he stood firm against the Union government‘s undue interference in Zanzibar affairs. He engineered the attempt of Zanzibar to join the OIC in 1992 as a way of expressing Zanzibar‘s autonomy. Incredibly, the Commonwealth-brokered first Muafaka in June 1999 came about during his term and promised major structural and policy reforms including commitment by both parties to

- 16 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

work together in the spirit of national reconciliation to consolidate democracy in Zanzibar, promote human rights and good governance and ensure that the elections scheduled for the year 2000 and all other subsequent elections were free of controversy and in which the will of the electorate will be respected (Amnesty International, 2000).

In short the first Muafaka was wide ranging, containing 15 items that included aspects of the ZEC, the constitution, the judiciary, the electoral laws, the state media organs, a permanent voters‘ register, the freedom of political parties, civic education, attendance in the House of Representatives, promotion of good governance and democratization (the East African, in Mpangala, 2006). Issues of domination by the ruling party and exclusion of the dominated party cropped up and the agreement contained provisions for finding an independent assessor to examine claims of destruction of property, unfair dismissals from government employment, the withdrawals of student scholarships andprovisions to allow allegations of human rights abuses to be taken to the courts. Those who claimed to be dismissed from government employment were people of Pemba origin thought to be followers of CUF. By the October 2000 election—which was supposed to be the benchmark for implementation of this agreement—not only had the agreement not been implemented, but the human rights situation was deteriorating. The 18 ―prisoners of conscience‖, as the Amnesty International came to refer to the CUF leaders who were arrested on Amour‘s order for alleged ―treason‖ trial, were still languishing in jail. Clearly Amour wanted to divert the little international attention that Zanzibar attracted to allow him a free hand in quashing the opposition.

Karume,6 on the other hand, came to power after the October 2000 election with the strong backing of CCM Mainland and was not as popular as his close intra- party competitor Dr Gharib Bilal in Zanzibar. Ostensibly Bilal came first in the CCM Zanzibar caucus voting only to be overturned at the CCM Central Committee on the Mainland. His election to the presidency revealed the tension between the Union and Zanzibar government and the level of resentment over ―undue interference of the Mainland in Zanzibar affairs‖ grew. The violence on January 26/27—the largest ever after the 1964 Revolution—did more harm than good in boosting his popularity in Zanzibar. By allowing the reform of ZEC to include Commissioners from both CCM and CUF after the second Muafaka in 2001, Karume was seen among CCM hardliners as being too liberal and giving too much to the CUF. So although the 2005 election was still contentious, the level of contestation from CUF declined. For instance, although CUF refused to recognize Karume as Zanzibar president, they still attended sessions of the House of Representatives. What was evident however was the fact that the Pemba-Unguja divide was crying out for more attention. Unless Karume strove for a sustained attempt at reconciliation the only legacy he

6Amani Karume is son of the leader of the 1964 Revolution and First President of Zanzibar. Karume was assassinated in 1972 in what is believed to be revenge related to the Revolution. His major legacy is to have enabled majority Africans to gain power in Zanzibar. It is also generally agreed that Zanzibar made greater stride in development during his term in office including the modernizing of housing that are still in use presently.

- 17 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

would leave behind would be the January Killing. For him negotiating a workable peace deal became an opportunity to leave behind a favourable legacy.7

It is not surprising then that when inter party negotiations had apparently stalled and were likely to follow in the previous pattern of failure, Karume made a surprise move to invite Hamad to the state house on 5 November 2009 to discuss the fate of peace in Zanzibar. This completely bypassed CCM formal organs. This move earned Karume two crucial credits.First he was seen as a new beacon of hope in Zanzibar. Secondly, and most importantly, he was asserting Zanzibar‘s autonomy to deal with its own affairs. It should be recalled that the inter party negotiations were seen to be driven by the Mainland and people in Zanzibar were not entirely happy about such a development. The fact that final decisions had to be endorsed by CCM Mainland irritated a considerable constituency in Zanzibar. CCM Mainland had interests that did not necessarily coincide with the general interest in Zanzibar although they might have misread the intention of Karume.That the Union government decided to keep a brave face was in no doubt due to its surprise in the manner in which Zanzibar took up the matter in its own hands.The matter has since remained within the domain of Zanzibar and is no longer decided by any Union organ either of the party or the government. Immediately after the talks the leader of the opposition camp in the House of Representatives AbubakarKhamis Bakaritabled a private motion to ―support‖ the talks in February 2010 and called for formulation of Government of National Unity (GNU). The House passed the motion. Two weeks later the Attorney General drafted a bill to allow for a referendum on power sharing be ready for tabling in the next session of the House of Representatives in March 2010.The bill was passed by majority vote.

The critical question at this pointrelates to why power sharing was accepted in 2010 while it has been floated since 1995 and sternly refused by the dominant CCM. Power sharing is no longer a novel strategy to resolve conflicts in deeply divided society but it is tricky since it places a lot of demands on the current regime to compromise. In many cases old guards in dominant parties would not allow power sharing simply because their own positions become jeopardized. Political development within the region must also have had an impact on the Zanzibar conflict. The recent cases in the last three years include Zimbabwe and Kenya. Both countries face similar problems to Zanzibar and both have adopted power sharing structures. In Kenya power sharing preceded constitutional reform while in Zanzibar constitutional reforms preceded the new government structure.

THE REFERENDUM AND THE ROAD TO THE GNU In July 2010 a majority (67%) voted in favour of the GNU in a referendum.8 The results could be interpreted as a reverse order of thegovernment‘s usual approval

7 Although there have been six presidents in Zanzibar, there is a feeling that very few left behind strong legacies to be remembered with and this is a strong motivation for presidential actions in Zanzibar. 8The referendum question asked roughly translated was: ―Do you agree that a new form of government will be formulated after the general election of 2010?‖

- 18 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

in Zanzibar.9 If referendum results are read side by side with the general election results and taken as a level of approval of the government in Zanzibar, the first impression is that while Pemba generally would disapprove of the government during the election, it highly approved government‘s decision to formulate the GNU, while Unguja disapproved of the government. The ―Yes‖ vote in Pemba was in the region of over 80% while in Unguja it was in the region of 50-55%.

The agreement leading into the GNU is simply premised upon a notion that the government formed after the 2010 election in October would include CUF and CCM in a power-sharing arrangement. The power-sharing formula agreed is straightforward. A winning party produces the president while the runner up produces first vice president. The second vice president comes from the president‘s party and becomes the leader of government business in the HoR. In addition Cabinet posts are distributed in proportion to the number of elected seats each party obtained from the election. In the 2010 election held on the 31st October 2010 CCM‘s presidential candidate came first winning 51.1% of popular votes while CUF‘s presidential candidate came second with 49.1% of popular vote (See Table 5 above). Accordingly, the CCM‘s candidate is now the president; CUF‘s becoming first vice president. With 22 of the 50 seats in the HoR CUF got 10 cabinet posts while CCM got 15. This makes a cabinet of 25, considerably inflated for a population of 1 million compared to the preceding cabinet which had 16 ministers and deputy ministers. But this is a cost of a power-sharing government built on consociational principles. Similar arrangements have been implemented in Kenya and Zimbabwe.

The referendum was very interesting in the way it was crafted. It was clear that CCM and CUF had taken positions and were simply trying to get popular support. Unlike Kenya, which had held a more similar referendum just a few days after the Zanzibar one, there were no open campaigns in the run up to the referendum. The government explained that allowing for open campaigns would politicize the process and dent people‘s choice. They maintained that people had to decide without any undue interference from the political parties. Some CSOs‘ got funding from UNDP to conduct voter education for duration of barely four weeks. Interestingly, much of the content provided by the voter education simply encouraged people to go out to vote ―Yes‖. Again, the president used various occasions to talk about the GNU. Through these rallies Karume‘s message was clear; there was a need to break new grounds for peace in Zanzibar and therefore people should vote ―Yes‖. CSOs carried a similar message in order to deter giving people an opportunity to sift through positive and negative elements of the GNU. It was portrayed as if the GNU was the only way forward for Zanzibar. To show his seriousness, Karume fired one of the Regional Commissioners in Unguja, who was thought to have been conducting a clandestine campaign against the government of National Unity. This drove further underground any ―No‖ voter crusaders while leaving the pro-Yes camp both in CCM and CUF a free ground to campaign.

9The author of this paper spent four weeks in Pemba as part of a local observer team, Tanzania Election Monitoring Committee (TEMCO). Much of the information in this section comes from direct observation of the process leading to the referendum.

- 19 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

CONCLUSION This paper has attempted to contextualize the politics of identities and how they impinge on the current effort in conflict resolution in Zanzibar. It has argued that while violence in Zanzibar is not comparable to other conflicts across the continent and beyond, it has most of the characteristics of a protracted and deep-seated ethnic conflict. The political divisions are superimposed on deeper racial/ ethnic divisions embedded in territorially-defined horizontal inequalities. The 1964 Revolution simply shifted power balances in racial relations but left the racial boundaries intact. It is hardly surprising therefore that when in 1992 after four decades of policies that purported to erase ethnic and racial politics, old boundaries resurfaced even more pronounced.While logically power mighthave been the most appropriate form of government, strong historical memories have kept ―African‖ majority and ―Arab‖ minority apart. The identity question is complicated since in Zanzibar it is not simply the identity of individuals that is at stake and prone to self re-categorization and labelling but also the identity of the state. The short-lived pre- Revolution government identified itself as an ―Arab‖ state and the quick bid to ―Arabize‖ the bureaucracy and the security forces considerably undermined defence and paved the way for the easy capture of government by ―African‖revolutionaries. The question after the 1964 Union has been ―Is Zanzibar a state?‖ The answer to that question has been subject to many debates and has cost some individuals political fortunes. Herein lies a puzzle—on the one hand the question of identity acts as a strong divisive force in Zanzibar while on the other it acts as force of unity. The current peace settlement therefore symbolizes the struggle both for individuals in Zanzibar to reclaim their identity and with it a fair share of power while at the same time searching for the lost glory of the state of Zanzibar.

This hypothesis explains why Maridhiano worked in 2010 whereas the first and second Muafaka failed. While the first and second Muafaka had a strong Mainland hand, Maridhiano, more than any other previous peace processes, offered an opportunity for Zanzibar to reassert its autonomy, however symbolic and nominal this might seem.It is definitely clear that the agreement was skilfully crafted such that CCM does not lose much in terms of its identity. Despite CUF‘s repudiation of the Revolution, the Revolutionary Government and the Revolutionary Council— powerful identity symbols in Zanzibar—these remain intact. In short CUF has been simply invited to serve in the Revolutionary Government. CUF on its part has reached what has been a very difficult achievement to make; it has taken Pembaback in the political map of Zanzibar.

REFERENCES Amnesty International (2000) Tanzania: Prisoners of Conscience Face Treason Trial in Zanzibar. AI Index: AFR 56/01/00

Amnesty International (2001) Tanzania: Inquiry into Zanzibar killings must be prompt, independent and impartial. AI Index AFR56/013/2001-News Service Nr.203

- 20 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

Anglin, Douglas G., (2000) Zanzibar: Political Impasse and Commonwealth Mediation; Journal of Contemporary African Studies Vol. 18, No. 1, pp. 39-66

Bakari, Mohamed A., (2001) Democratization Process in Zanzibar: A Retarded Transition; Institute for African Affairs, Hamburg

Bakari, Mohamed A., (2007) Understanding Obstacles to Political Reconciliation in Zanzibar: Actors, Interests and Strategies; Paper Presented at the Methodology Workshop Organized by the Concern for Development Initiative in Africa (ForDIA), Belinda Ocean Resort, Dar es Salaam, 22-24 June

Baregu, Mwesiga (2007) Understanding Obstacles to Peace in the Great Lakes Region: Actors, Interests, and Strategies; A Conceptual Framework Paper and Research Proposal Submitted to Peace, Conflict and Development Program Initiative International Development Centre Eastern and Southern African Regional Office, Nairobi, Kenya

Cameron, Greg (2002) Zanzibar’s Turbulent Transition; Review of African Political Economy, Vol. 29, No. 92, June, pp 313-330

Campbell, Jane (1962) Multiracialism and Politics in Zanzibar; Political Science Quarterly, Vol. 77, No. 1, March, pp 72-87

Daily News (Tanzanian Newspaper) 26th May 2008

Daly, Samuel (2009) ―Our Mother is Afro Shirazi and Our Father is the Revolution‖: The 1964 in History; Unpublished Senior Thesis, Columbia University Department of History

Fearon, James D., (1999) ―Why Ethnic Politics and ‗Pork‘ Tend to go Together‖; Paper Presented at a MacArthur Foundations-sponsored Conference on Ethnic Politics and Democratic Stability held at Wilder House, University of Chicago, May 21-23

FEWER (2005) Zanzibar Electoral Violence and Reconciliation; A Report of the Forum on Early Warning and Early Response-Africa (FEWER), Nairobi, Kenya

Groot, Emile de (1953) Great Britain and Germany in Zanzibar: Consul Holmwood’s Papers 1886-1887; The Journal of Modern History, Vol. 25, No. 2, June, pp 120-138

International Federation for Human Rights and Legal and Human Rights Centre (2001) A Fact Finding Report on Police Brutality and Election Mismanagement in Zanzibar, Report No. 307

Killian, Bernadeta (2008) The State and Identity Politics in Zanzibar: Challenges to Democratic Consolidation in Tanzania African Identities, Vol. 6, No. 2

Lofchie, Michael (1963) Party Conflict in Zanzibar; The Journal of Modern African Studies, Vol. 1, No. 2, June, pp 185-207

Mamdani, Mahmood (2001) When Victims Become Killers: Colonialism, Nativism and the Genocide in Rwanda; New Jersey, Princeton University Press

- 21 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

Mbunda, Richard (2009) Resolution or Containment? An Evaluation of Direct Negotiation and Settlement Facilitation in the Zanzibar Conflict; Department of Political Science, University of Dar es Salaam

Mpangala, Gaudens Phillip (2006) The Zanzibar Conflict: A Search for Durable Solutions; in Rupia, Martin; Lwehabura, Jonathan and Roux, Len (eds) Civil Security Relations in Tanzania: Investigating the Relationship Between the State, Security Services and Civil Society; the Institute for Security Studies, Monograph Number 18

Mukangara, Daudi R., (2000) Race, Ethnicity, Religion and Politics in Zanzibar; in Maliyamkono, Ted L., (ed) The Political Plight of Zanzibar: EUSARP, Dar es Salaam

Mwananchi Newspaper, Tanzania, 29th March 2010

Ndumbaro, Laurean (2003) The State of Constitutionalism in Tanzania; Unpublished paper, REDET, Department of Political Science and Public Administration, University of Dar es Salaam

Newbury, Catharine M., (1983) Colonialism, Ethnicity and Rural Political Protest: Rwanda and Zanzibar in Comparative Perspective; Comparative Politics, Vol. 15, No. 3, April, pp. 253-280

Papagianni, Katia (2008) Power Sharing Transnational Governments and the Role of Mediation; Centre for Humanitarian Dialogue, The Oslo Forum Network of Mediators

Pottie, David (2002) Party Strife and Political Impasse in Zanzibar’s October 2000 Elections Representation, Vol. 38, No. 4, pp 340-350

RaiaMwemaNewspaper, Tanzania, 3rd – 9th March, 2010

Rawlence, Ben (2005) Briefing: The Zanzibar Election: African Affairs Vol. 104, No. 416 pp 515-523

Revolutionary Government of Zanzibar (2004) Household Budget Survey; Office of the Chief Government Statician

Rogier, Emerick (2004) Rethinking Conflict Resolution in Africa: Lessons From the Democratic Republic of the Congo, Sierra Leone and Sudan; Netherlands Institute of International Relations, Conflict Research Unit

Rwambali, Faustine (2001) Zanzibar’s Long History of Political Violence; The East African, 5th February

Sheriff, Abdul (2001)Race and Class in the Politics of Zanzibar: Africa Spectrum, Vol. 36, No. 3, pp. 301-318

Tambila, Kapepwa I., (2000) Aspects of the Political Economy of Unguja and Pemba; in Maliyamkono, Ted L., (ed) The Political Plight of Zanzibar: EUSARP, Dar es Salaam

Todd, Jennifer (2005) Northern Ireland: The Changing Structure of Conflict in Coakley, John, Brigid, Lafan and Todd, Jennifer, Renovations or Revolutions? New Territorial Politics and the United Kingdom, Dublin, University College Dublin Press

- 22 - IBIS DISCUSSION PAPERS PATTERNS OF CONFLICT RESOLUTION

URT (1999) Political Service Retirement Benefits Act

URT (2003) The 2002 Population and Household Census; National Bureau of Statistics, Dar es Salaam

Varshney, Ashutosh (2001) Ethnic Conflict and Civil Society: India and Beyond World Politics, Vol. 53, No. 3, April, pp. 362-398

Zanzibar Election Commission (2010) The 2010 Presidential and Representatives Results, www.zec.go.tz (1st December 2010)

- 23 -