Race and Class in the Politics of Zanzibar Author(s): Abdul Sheriff Source: Africa Spectrum, Vol. 36, No.

3 (2001), pp. 301-318 Published by: Institute of African Affairs at GIGA, Hamburg/Germany Stable URL: http://www.jstor.org/stable/40174901 Accessed: 26/10/2010 20:51
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=iaagiga. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Institute of African Affairs at GIGA, Hamburg/Germany is collaborating with JSTOR to digitize, preserve and extend access to Africa Spectrum.

http://www.jstor.org

afrikaspectrum 36 (2001) 3: 301 - 318

__^

Abdul Sheriff

Race and class in the politics of Zanzibar

The prevailing interpretation of the recent history and politics of Zanzibar has been based on the contention that race and class coincided on the isles. The 1964 revolution was therefore seen as the overthrow of an Arab landowning minority by the African majority. (Lofchie 1965. Mrina & Mattoke n.d. Mapuri 1996) The massacre and expulsion of a large number of Arabs, and the nationalisation and redistribution of large landholdings following the revolution should therefore have solved the problem. However, the history of post-revolution Zanzibar has been punctuated by repeated recurrence of political unrest and a persistence of deep political division. Although the hackneyed racial labels are still touted to explain the political divide, it to see the is clear that the pre-revolution social structure has to be re-examined time. The social at that and even race and class fit between of political degree fault-lines in the society can be traced back to the history of Zanzibar before colonialism and even more to the period of colonial rule. These contradictions were not allowed to be played out by the colonial authorities, and they were not properly healed after the Revolution; on the contrary, in some ways they have been exacerbated.

Race & class in Zanzibar

During the nineteenth century Zanzibar had developed as the capital of a Commercial Empire based on the twin foundation of commerce and a plantation economy on Zanzibar. The first sector had centralized the foreign trade of much of eastern Africa on Zanzibar, and it gave rise to a prosperous mercantile state and a merchant class which was notably though not entirely Indian, and the beginnings of an urban working class that was partly servile, handling the commercial economy. The plantation economy, on the other hand, created a slave society with rich landowners many though not all of whom were Arabs, and slaves imported from the mainland. This gave rise to a racial paradigm during the colonial period that tended to label population by race, and race denoted function: thus Arabs were landownthe downtrodden. as Africans the and were Indians merchants, ers, these ethnic that none of the fact characterisation a Such ignored (Flint:1 965:651) categories were undifferentiated, and more importantly, they were undergoing very fundamental economic and social transformations triggered by the imposition of colonial rule. social structure Even before the imposition of colonial rule the racial-cum-class was by no means clear-cut. To start with, the Hadhrami Arabs did not form part of the ruling class, and many had migrated to Zanzibar to work as porters at the busy 301

Abdul Sheriff

commercial centre or to open small shops; and even among the Omanis, a large number probably migrated as poor Mangas (as recent Omani immigrants are called in Kiswahili), hoping to improve their lives as small shopkeepers in the countryside, as they continued to do until the 1964 revolution. While a small percentage of the Indians were rich merchants, and some of them were fabulously rich, the majority had migrated from a poor region in India and had established themselves as small shopkeepers or clerks in the larger Indian firms with the hope of setting up their own small shops eventually, while some remained poor washer-men right down to the present. Finally, despite the tendency to lump all Africans as belonging to a single class, there were significant and portentous differences between on the one hand the free indigenous people who were never enslaved, and some of whom even owned large plantations and slaves, and the slaves who had been imported in large numbers during the nineteenth century to work on the clove plantations. The colonial partition of the Commercial Empire towards the end of the nineteenth century was to add to the fuzziness of the ethno-class divisions of the Zanzibari society, and in some ways to transform it. In the commercial sector, the partition of Zanzibar's commercial hinterland diminished the entrepot role of Zanzibar that sealed the fate of the merchant class and it began to disintegrate. As the colonial powers on the mainland diverted the trade of their new colonies to their own metropoles to Hamburg and London, Zanzibar was bypassed and was gradually reduced to a small island economy. With a population of less than 200,000 that was moreover getting impoverished, many of the Indian merchants began to migrate to the mainland. Those who remained on the islands had to be content with the small internal market, and some turned to money-lending to the landowners. During the Depression of the 1930s the plight of Zanzibar's economy and that of its financiers became all too apparent. In an effort to cut out the middlemen the British colonial authorities sought to eliminate the Indian merchants and moneylenders by establishing the Clove Growers' Association (CGA) to finance and market Zanzibar's cloves in collaboration with British banking capital. With their back to the wall the Indian merchant class fought back with a boycott of the clove trade. They were supported by the Indian National Congress to reach a compromise to retain some share of the clove trade. Henceforth, they could play only second fiddle in the politics of Zanzibar. The mainstay of Zanzibar's economy during the colonial period was therefore clove production. However, the abolition of the slave trade in 1873 and the abolition of slavery in 1897 began to transform the formerly slave-based plantation economy into a colonial capitalist economy based on squatter and wage labour. This destroyed the economic base of many of the large plantation owners who were impoverished, forcing them to sell at least part of their land, reducing many of them to little more than peasants. Sons of many landowners had to seek employment in the civil service as teachers, clerks, etc. (See AN Muhsin n.d.) The decline of the landlords was so serious that some colonial officials in the 1920s began to question whether they were not a spent force and should be jettisoned in favour of peasants who were considered 'the cheapest instrument' of colonial production. (Flint 1965:655) However, the conservative colonial regime was too timid to contemplate a social revolution that such a policy would have entailed, and in the

302

Race and class in the politics of Zanzibar

1930s it intervened to bail out the landowners with debt relief and the formation of the CGA, thus hampering a painful but in the long run perhaps a more gradual and peaceful resolution of the contradictions in the economy and society. The decline of the big landowners provided an opportunity for the growth of an indigenous peasant class, especially in Pemba. They bought small parcels of land from the heavily indebted Arab landowners, or planted their own lands with cloves. Their way of life and economic fortunes differed little from a large section of the declining Arab landowning class many of whose members were becoming peasantised. They thus developed common interests and political outlook that cut across ethnic lines.

The genesis of a regional divide

The socio-economic transformation, however, was not uniform across the two islands, and this was going to be significant in the future political history of Zanzibar. The larger Unguja island was the first place where cloves were planted on a largescale from the 1830s, but the steep decline in the price of cloves by the late 1840s had hampered its spread to Pemba until the hurricane of 1872 which wiped out most of the clove trees on Unguja. By then the relative poverty of Unguja's soils and the heavier rainfall and richer virgin forests of Pemba that were even better suited for cloves had become apparent. However, the following year all slave trade by sea was banned. Thus, while some of the landowners tried to transfer their slaves from Unguja to Pemba, or smuggle them from the mainland, there was an acute shortage of labour there. This forced them to come to an arrangement with the local peasants to help establish new clove plantations. As peasants not yet fully integrated into the money economy, they were less interested in a cash wage than in getting clove lands for themselves under the so-called 'half-and-half system by which the labourers were given half the cleared and planted land. They therefore began to acquire their own small clove plantations or began to plant cloves on their own lands. The historical differences in the patterns of development of landownership between the two islands emerge very clearly from the graphs drawn from registers of the Clove Bonus Scheme of the 1920s. In Unguja only 2.4 per cent of all clove owners (owning more than 1 ,000 clove trees) owned 40 per cent of the clove trees, and 83 per cent of these largest landowners were Arabs. At the bottom of the pyramid 78 per cent were small landowners (owning less than 100 trees), and they owned 18 per cent of all the clove trees between them. A vast majority of them were indigenous peasants (52 per cent) and recent African immigrants from the mainland (16 per cent). While the class structure is very sharply defined and the ethnic factor is more marked, it is nevertheless important to note that the Arabs (8 per cent) also constituted a part of this poorest class. Moreover, looking at the Arabs as an ethnic group, only 10 per cent were big landowners and 79 per cent were poor or middle peasants even in Unguja in terms of clove ownership, (see Graph 1)

303

Abdul Sheriff

Graph 1/a: Race & class: Landownershlp

In Zanzibar. 1922-29

Pemba

1000+

\^-

- **** Hi

500-999

J^>

V.

100-499
% 30

^HiiH^H8V^^MfflH^
20 10 0 10 20 30 40 50 60 70

% of clove owners / % of clove trees No.ofCI. I

trees tl1 ers T^ per Ug^lgfl^Arg^ J^jgJ,^-^1, Jf\f ^con[Q

1000+

0.4/ / /3.1

1.8/
/

18.9/

0.2/ / 1.4

0.04/
/

0.02/
/

24/
/

/0.5

/0.2

/26.1

300-999 y//
/l.9 / 100-499 14/18.5/ ////// 1.8 /

2y^

10.8

y^2.3 17.8/

2/
2.9/ / 7.3

y^ 0.1

/
y^O.2 0.6/412/

4 y^
/13.5

22.6

17.8 23.6/

2.7 y/ >/i.9

0.7

/45s 52.1/^ yi2.9 \&* /

It /^O^ Total

MAy/ y44 40.4 /^

0.7 >/ /)i \A/

/"fe.1 4*2/

33/^ //////

10.5./

^/9.1

^56.7

^/27.8

y^5.1

13

X 100

304

Race and class in the politics of Zanzibar

Graph 1/b: Race & class: Landownership

in Zanzibar, 1922-29

Unguja

40

30

20

10

10

20

30

40

50

60

% of Clove owners / % of Clove trees

y

\r*esper llndiansl Arabs iShtrazi [African Others Total

of Cl fill

500-999

02/ / Xi.3 0.7 /^ ////// /i.3 t.4/ y^O.5

2\ / / -/ 11.0 7^>/ Xi3.0 84y/ >/2.3 19.7/ / 60.3

0.3/ / /l.8 6.1/ X 8.2 31.6/ /12.1 58.1/ X 232

0.1/ / X0.3 t4/ /2.1 13 6/ z^3-1 17.1/ X 5.8

0.1/ / /0.6 O"7/^ /i.O \& /0 2.3/ /3.8 / 4

2.8/ / /15.0 16.6/ /2S.9 785/ ^18.6 100.3/ XiOO.1

100-49$

Total

2.3/ ////// /6.8

Source: Zanzibar Archives: Clove Bonus Scheme Registers, 1922-29.

305

Abdul Sheriff

It is this pattern of landownership that was used by Lofchie as the general pattern for Zanzibar as a whole, completely misreading his source that had specifically stated that it referred only to Zanzibar island (Unguja), and that corresponding figures for Pemba were not then available. (Lofchie 1965:85. See Pirn 1932:83) However, only 20 per cent of the clove trees were in Unguja, and the major producer of cloves was Pemba. In that island the largest landowners coincidentally constituted exactly the same 2.4 per cent of all the clove owners, but they owned only 26 per cent of the clove trees between them; of these three quarters were Arabs. However, what is marked about Pemba is that poor and middle peasants, who constituted 93 per cent of all the landowners, owned 58 per cent of the clove trees between them. Moreover, in sharp contrast to Unguja, these two strata were much more symmetrical in terms of ethnic groups with almost as many Arabs as indigenous Wapemba peasants. Looking at the Arabs as an ethnic group less than five per cent of them were big landowners and 88 per cent were middle or poor peasants. To get an ethnic and class structure applicable to Zanzibar as a whole, one would need to combine the two graphs, taking into consideration the much greater weight of the Pemba figures, and the final graph would be much closer to the Pemba pattern than to the one for Unguja. But our interest is not as much to get the general picture as to get the significant differences between the two islands in order to understand their different political behaviour before and since the revolution. It is obvious from the graphs, and especially for Pemba, that race and class did not coincide, but rather that the ethnic divide cut right across class lines. The actual life condition of the middle and poor peasants, whether Arab or indigenous Zanzibaris, could not have been different under conditions of economic depression during the colonial period, and especially during the Great Depression and in the postKorean War bust. This common condition created there a different dynamic of racial intercourse from that painted by Lofchie and others. This was reflected in a greater degree of intermarriage, (Tominaga & Sheriff 1990) and ultimately in political collaboration and solidarity before the revolution and since. Lofchie had correctly argued that in a mono-crop agricultural colonial economy like Zanzibar, landownership was a major indicator of social distribution of wealth. However, he failed to explain the genesis of a significant difference in the social and political behaviour of the two islands. Moreover, he ignored the significant differences between the indigenous people of Zanzibar and the more recent immigrants from the mainland, treating them all as an undifferentiated African majority. Lofchie's interpretation of an African revolution was very popular in the 1960s, the decade of African independence, and it has continued to form the basis of most of the subsequent secondary analyses of Zanzibar's politics. Lofchie's table of landownership was picked up by the ideologues of the Zanzibar revolution to construct a mythology of the ruling party in Zanzibar to justify the revolution and all that followed. Thus, Mrina and Mattoke of the CCM Party Ideological College, who wrote Mapambano ya Ukombozi Zanzibar, (The Struggle for the Liberation of Zanzibar) in Kiswahili for use in all schools in Tanzania, reproduced Lofchie's table. Unfortunately, they unintentionally misplaced the figures of

306

Race and class in the politics of Zanzibar

the smallest landowners between Arabs and Africans, thus ended up showing the Arabs were the largest number even in this category, although the totals and percentages are still as given by Lofchie. (Mrina & Mattoke n.d.:36) Mapuri, the former Deputy Chief Minister of Zanzibar, who wrote The 1964 Revolution: Achievements and Prospects in the immediate aftermath of the 1995 multi-party election to justify the disputed victory of the ruling party, went on to 'correct' Mrina and Mattoke's mistake by merely adding up the totals, including the misplaced figures, without even consulting Lofchie. Thus, according to his tertiary account, 79.9 per cent of all the landowners of all classes were Arabs, whereas Lofchie's original table would have given only 26.9 per cent. (Mapuri 1996:91) Neither of these authors cared to check the original source where they might have been able to correct Lofchie's mistake instead of compounding it with their own; they merely copied each other's tables to prove their pre-determined position.

The political divide

The critical test for the role of ethnicity is in the political behaviour of the various social groups. Ethnicity is not a biological category but a sociological dimension. It refers to group consciousness based on perceived common cultural characteristics, such as language, religion, traditions, social behaviour and a sense of its history. With a cosmopolitan population that has shared a common language, religion and a common history over many centuries, these categories were not frozen in Zanzibar. Between 1924 and 1948, for example, the Swahili who had numbered 34,000 in 1924 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 1931 , before they also apparently committed an ethnic harakiri by 1948, although in this case it was because the colonial authorities preferred tribal rather than the broader 'Swahili' and 'Shirazi' categories. Even the mainland Africans, who had numbered 65,000 in 1924, seem to have diminished to 51,000 by 1948 as some of them were indigenised. At the same time, the indigenous 'tribes' of Zanzibar, the Hadimu, Tumbatu and Pemba, and even the Arabs, grew enormously between 1924 and 1948. (Zanzibar Protectorate, 1924, 1931, 1948) What is clear is that the so-called ethnic boundaries were extremely porous and spongy, allowing constant assimilation and re-identification of people as historical conditions demanded. Moreover, there were frequent intermarriages between these ethnic categories. An analysis of birth certificates between 1953 and 1963 shows that 9 per cent of the marriages cut across ethnic or 'tribal' boundaries. Of children born among the Hadimu of Unguja, 39 per cent had one parent from another 'tribal' or ethnic group, 22 per cent among the Shirazi, and 10 per cent among the more isolated Tumbatu. Intermarriage was particularly common between the Shirazi and the Arabs in Pemba, and between the Hadimu and mainland Africans in Zanzibar town. (Tominaga & Sheriff 1990)

Abdul Sheriff

However, the colonial rulers preferred to deal with definable and localised groups in the time-proven policy of divide and rule, and this had the effect of strengthening and freezing communal divisions. In a cosmopolitan society like Zanzibar, there was a plethora of ethnic and communal associations through which the British preferred to deal with the local population, each dominated by its own elite. had to go through a double transformation before itThe African Association could claim to be truly representative of all the 'African' population of Zanzibar. The first was the Arab Association which was formed early in the twentieth century to protect the interests of the declining landowning class. By recognising the Arab Association as the only legitimate representative of the Arabs or of landowning interests, the colonial rulers tended to ignore the differentiation that the community was going through as well as other (non-Arab) landowning interests. Not surprisingly, the second to be formed was the Indian National Association that was set up precisely to struggle against the declining fortunes of the merchant class. It had its political moment in the 1930s when it was forced to put up a fight against total exclusion from the clove economy during the Depression. They won a reprieve in the of 1938, but the struggle showed that the merchant class was too compromise small to act independently in the political arena. On the other hand, the African Association was formed in the mid-1 930s in response to the hardships of the Depression felt by the urban working population, most of whom were Africans of mainland origin. Not surprisingly, it had close links with the African Association in Tanganyika, and its leadership was initially taken by the Christian mission-educated elite, but a majority of the population in Zanzibar, even those of mainland origin, was Muslim. There was therefore a cleavage that was both ethnic and religious between the African Association and the indigenous population of Zanzibar. As late as the mid 1940s the African Association was arguing, In the name of Democracy, Christianity and human equality', that the Shirazis were Asians, and therefore could not represent the interests of the Africans of Zanzibar. (Mrina & Mattoke n.d.:51-2) The first was a colonial ruling that civil servants could not engage in political activities. (Lofchie 1965:163-5) This was directed primarily at the Arab Association that was then challenging British authority in Zanzibar, but the landowners had an independent economic base, and they were not unduly affected. In the case of the African Association, it has been interpreted by post-revolution nationalist historians as a cynical colonial manoeuvre to deprive it of an educated leadership, since most of the educated Africans were civil servants. (Mrina & Mattoke n.d.:54) NevertheChristian educated the move less, by eliminating a predominantly leadership, brought the African Association closer to the African population that was Muslim and not so well educated. Its future leaders were going to be much more earthy. did not overcome This move, however, the ethnic divide between the mainlanders and the indigenous population of Zanzibar, the Shirazi. The former consisted of descendants of freed slaves and recent immigrants from the mainland, some of them living as squatters on the plantations but a majority were urban workers in Zanzibar town. The Shirazi, on the other hand, were predominantly peasants, and it is not surprising that they were the last to articulate their interests organisationally. They were not a uniform sack of potatoes but a class that was go-

Race and class in the politics of Zanzibar

ing through a process of social differentiation as a result of their incorporation into the world capitalist economy. (Sheriff 1991) By the 1920s there was a sizeable of their class of middle and rich peasants among them, and they were conscious different interests from those of the urban working population of mainland origin. In 1940 they formed the Shirazi Association in Pemba, to preserve 'all that is best in as a 'foreign association1 country life'. They described the African Association which did not represent the 'the natives of this island'. Although critical of domination by large Arab landowners in the articulation of landed interests, the Shirazi peasants shared common economic interests with their Arab counterparts as clove producers, a majority of whom, as we have seen, had been virtually peasantised, and with whom they lived cheek by jawl. (Tominaga & Sheriff) But the middle and rich peasants of Pemba were only one product of that process of differentiation; the other was the poor peasantry, located largely in the eastern and southern parts of Unguja and to some extent in eastern Pemba, outside in the rocky inthe clove plantation areas. Unable to sustain their self-sufficiency forced to were sell their labour environment there, many poor peasants hospitable seasonally to pick cloves on plantations owned by both Arab and Shirazi landownor to migrate to Zanzibar town where they mingled more ers and rich peasants, the with working population of mainland origin. Although their leaders asked freely in 1941 (Zanzibar Archives: AB 12/2), the Shirazi to join the Shirazi Association and economic interests poor peasants of Unguja had more common experiences with the urban mainland African population with whom they were eventually to form a political alliance in the 1950s. but in all there were nearly two These were the four main ethnic associations, dozens which were recognised and nurtured by the colonial state. (Babu 1991:224) It tried to freeze the Zanzibari society in an ethnic mould, and by selecting the leadit reinforced for their 'communities,' ing landowners or merchants as spokesmen It communities. leadership of small elites over these socially quite heterogeneous of appointed big Arab and Indian landowners and merchants as 'representatives' their communities to the legislative Council, and from the mid-1 940s, a Shirazi to represent the 'Africans'. However, this was bitterly resented by the mainland Africans who said the Shirazis were Asians and could not represent the interests of must have the Africans of Zanzibar. The Shirazi responded that any representative been resident in Zanzibar for five to six generations, hitting at the shallower roots of the mainland Africans in Zanzibar. (Mrina & Mattoke 1980:51-2) However, after the Second World War the ethnic mould was no longer tenable. It The first to challenge the colonial status quo was in fact the Arab Association. had previously enjoyed the most influence, but by the 1950s it had begun to realise that in a colonial society power rested decidedly with the colonial state, and that their so-called influence was an illusion. The more radical among their younger members realised that the interests of the more integrated Zanzibar society could no longer be served by communal associations. Although the Arabs constituted an a non-racial common roll election. demanded Association Arab the ethnic minority, This was a unique phenomenon only if one looks at Zanzibar through an ethnic The radicals went farther and decided to abandon ethnic politics altospectacle. gether. They joined a political party, the Zanzibar Nationalist Party (ZNP), which

309

Abdul Sheriff

was formed by some Shirazi peasants in Unguja. They were protesting against colonial regulations against anthrax that had led to the first political confrontation in 1952. It was a gamble on a national over an ethnic identity. It seemed to have failed miserably in the first election in 1957 when the ZNP failed to win a single seat. But an ethnic spectacle is misleading because it refuses to focus on social differentiation within these ethnic categories and their histories over the past century. This was to have a profound influence on political behaviour in Zanzibar during the Zama za Siasa (period of politics), as the period between 1957 and 1964 has come to be known. While many of the landowners and merchants had been impoverished and reduced to peasants and small shopkeepers, they found common interest with sections of the peasantry that was going through a process of social differentiation. Thus the ZNP eventually came to represent the interests of not only the landowners and peasants, especially in Pemba, but also shopkeepers and civil servants, united in a broad nationalist coalition. Even a section of the indigenous working class was attached to it, albeit precariously. The more radical youths of the party eventually broke away on the eve of the 1963 election to form the Umma Party, carrying with them the workers' federation. On the other hand, with the active encouragement of Tanganyika's president Julius Nyerere, the African Association combined on the eve of the 1957 elections with the Shirazi Association to form the Afro-Shirazi Party (ASP). While the bond between African working class and the Shirazi poor peasantry in Unguja proved durable, the Shirazi Association of Pemba was a very reluctant partner. In 1957 it set up its own independent candidates, and merged with the ASP only after the election. The partnership collapsed in 1959 when they broke away to form the Zanzibar & Pemba People's Party (ZPPP), essentially representing peasant interests in Pemba. In 1961 the ZPPP and ZNP formed an electoral alliance that survived through two elections to form the first independence government in 1963. The first elections on a limited franchise were held in 1957 to select six unofficial representatives to the Legislative Council. While the Zanzibar Stone Town seat was won by an Indian candidate of the Muslim Association, the ASP won three seats in Unguja and independent Shirazi candidates won the two seats in Pemba and they later joined the ASP. Whereas the ASP and its allies won a total of 22,024 (64 per cent of the total) and grabbed five of the six seats, the ZNP collected 10,982 votes (31 per cent) but did not win a single seat. This was a severe blow for the ZNP and for non-ethnic nationalism, but it proved to be wake-up call to reorganise the party and start serious mass mobilisation under its youthful new secretary general Abdulrehman Babu.

310

Race and class in the politics of Zanzibar

Table 1: 1957 Election results Party

ASP Sh.A. ZNP Ind. Total

Pemba Unguja I Votes I Seats I Votes

12.787~ 3 ~ 657 9.237 4.670 1.075 | 15.639

Total I Seats I Votes

0 2 0 0 \2

I Seats
3 2 0 1 6

6.312 0 1 623 | 19.722 | 4 (Source: Lofchie 1965:176)

13.444 (38%)~ 9.237 (26%) 10,982(31%) 1.698 (5%) | 35.361(100%) |

On the other hand, during the same period the ASP failed to maintain its shaky coalition with a majority of its Shirazi constituents, especially in Pemba. Its two members there as well as one from Unguja broke away in 1959 to form the Zanzibar & Pemba Peoples' Party (ZPPP). The new re-alignment was expressed in the January 1961 election when the franchise was broadened by the removal of some of the economic and educational qualifications. The two major parties emerged almost equally balanced, the ASP winning 43 per cent of the total votes and 10 seats, with its main base in Unguja, and the ZNP collecting 39 per cent of the votes and nine seats, almost equally balanced between the two islands in terms of votes and seats. Table 2: January 1961 Election Results Pemba Unguja Party I Votes
ASP ZPPP ZNP Total ~

Total I Seats I Votes I Seats

I Seats I Votes

8 27.495 0 1.946 5 15.555 I 44.996 | 13 (Source: Lofchie 1965:200)

~9.203 13.595 17.169 [ 39.967

2 3 4 9

36.698 (43%) ~^IO 3 15.541(18%) 9 32.724(39%) | 84.963(100%) | 22

The ZPPP won 18 per cent of the votes and three seats, almost entirely based in Pemba. It thus emerged holding a balance of power, but the political dust had not yet settled down. Under the strain of negotiations to form a coalition government, the ZPPP split with one member joining the ASP and the other two joining the ZNP. This created a tie, and an interim caretaker government under a British civil servant had to be formed to organise another election in June in which a new seat was added. In the meantime the ZNP/ZPPP formed an electoral alliance and won a parliamentary majority to form an internal self-government. Political competition was so intense that widespread rioting followed the June 1961 election in which 69 people were killed, mostly poor Manga shopkeepers and their families who were isolated in the countryside, setting the stage for political violence in the future.

311

Abdul Sheriff

The results of the January 1961 election exposed the limitations of an ethnic explanation of Zanzibar politics. The 1948 census which had identified ethnic divisions of the isalnds' population showed that the Arabs constituted only 17 per cent of the total population, but the allegedly Arab party, the ZNP, had received 39 per cent of the votes in 1961 , more than double that proportion. On the other hand, the mainland Africans constituted only 19 per cent of the total population, but it collected 43 per cent of the votes. In between the two ethnic extremes were the indigenous Shirazis who constituted a total of 56 per cent of the population. However, under the tensions of political competition the Shirazi did not behave as a homogeneous political entity and split three-ways politically. One portion supported the ZNP especially in Pemba and northern Unguja; another sided with the ASP in Unguja and southern Pemba; and the ZPPP grabbed the rest, primarily in Pemba. The final pre-independence elections were held in 1963 when universal adult franchise almost doubled the electorate, and the number of seats was increased to 31 . The ZNP/ZPPP electoral alliance proved durable, winning 46 per cent of the total votes in Zanzibar as a whole and 18 parliamentary seats. It had a strong base in Pemba where it garnered 56 per cent of the votes but it also had won a credible 36% of the votes in Unguja, primarilyin the north and in Zanzibar town. On the other hand, the ASP won 54 per cent of the votes in Zanzibar as a whole but only 13 parliamentary seats. It had its broad base in Unguja where it swept 63% of the votes and increased its share of the votes in Pemba from 37 per cent in June 1961 to 44 per cent. However, whereas the two parties had received their fair proportion of the seats in Unguja according to the strength of their votes, the ASP failed to translate their voting strength in Pemba into parliamentary seats. This was primarilydue to the fact that its support was more thinly spread throughout the island where poor peasants were probably more widespread in the plantations areas as well as in the poor coral bushland in the eastern third of the island. It therefore won in two constituencies by huge margins in the extreme north and south, but lost in four others by less than four per cent. The ZNP/ZPPP electoral alliance, on the other hand, was able to capitalise on their unity to sweep the parliamentary seats by clear but smaller margins. Table 3: January 1963 Election Results

Unguja
Party
"ASP ZPPP ZNP Total

Pemba
I Votes
33.853 21.037 21.376 76.268

Total
I Seats I Votes
2 6 6 14 87.085 (54%) 25.609 (16%) 47.950 (30%) 164.644

I Votes I Seats
53.232 4.572 26.572 84.376 11 0 6 17

I Seats
13 6 12 31

|__

| (100%)I

(Source: Lofchie 1965:218)

312

Race and class in the politics of Zanzibar

The only conclusion that can reasonably be drawn from these data is that while ethnic considerations were not irrelevant, the primary contradiction in the Zanzibari society on the eve of independence was neither race nor class exclusively, but a complex combination of the two, influenced by a national feeling in a cosmopolitan society. However, that national feeling was not enough to overcome the deep social divide born of a long and intricate history of social transformation. On the eve of independence from Britain in December, 1963, while the ZNP had formed an electoral alliance with the ZPPP which was a breakaway from the ASP, the ASP began to move closer to the Umma Party which was a breakaway from the ZNP. Whereas in 1957 the ASP and its Shirazi Association allies had won a total of 64 per cent of the total votes, its share of the vote in fact declined to 43 per cent in 1961 when it lost its Pemba Shirazi allies. With the broadening of the franchise and the split in the ZPPP in 1961, the ASP improved its electoral strength by winning back some of the Pemba Shirazi support, and it gained 54 per cent of the votes in 1963. The ZNP, on the other hand, demonstrated its success in popularising its nationalist over ethnic loyalties, and it increased its share of the votes from 31 per cent in 1957 to 39 per cent in 1961. However, while the ZNP had won over the ZPPP into a stable alliance, both were adversely affected by the broadening of the franchise that had the effect of reducing their total share of the votes to 46 per cent in 1963. 1957-1963

Table 4: Party strength in Zanzibar elections, I 1957 I 1957 Votes % Seats Party 3 38 ASP 2 Sh.Ass./ZPPP~~ 26 0 31 ZNP 1 5 Ind.

1963 I l963~ I 1/1961 I 1/1961 I Votes % Seats Votes % Seats 54 13 10 43 16 6 3 18 12 30 9 39

Total

I 100

| 100

I 22

| 100

[31

Source: Lofchie 1965: 176, 200, 218.

The revolution
The revolution of 1964 can best be understood not as an overthrow of an Arab oligarchy by an African majority as is commonly claimed, but as a civil war in a society that could not find enough common national ground. To explain the ethnic-cumclass content of the Zanzibar revolution one has to focus more narrowly on Unguja where class differences were sharper and the ethnic tendency was stronger. The clove plantations were concentrated in the western third of the island where cloves were first established in the nineteenth century using slave labour.

313

Abdul Sheriff

After the abolition of slavery some of the freed slaves remained on the plantations as squatters while others left for the town where they eventually formed part of the proletariat. The clove plantations had also attracted free migrant labour from the mainland during the twentieth century, some of them settling down on the plantations as squatters. Moreover, Unguja has a much larger area of coral bush in the east and south, inhabited by poor peasants who annually went to pick cloves in the plantation area, exposing them to the harsh conditions of labour migration. All of these sections of the population, including the landowners, depended on cloves that is notoriously unstable as a cash crop. The crisis sharpened in 1950s. During the Korean war the price of cloves shot up to unheard of levels, but within a couple of years it had collapsed to less than 10 per cent with the economic crisis in Indonesia, the main market. Landowners and squatters competed for the use of the same land for food crops to cushion themselves, and some squatters were evicted. Clove pickers from the coral areas were offered low picking rates, and the economic crisis had a direct effect on the well-being of the urban proletariat dependent on the commercial economy. With the skewed class structure in Unguja, the conflict appeared to be more racial, and it exploded in the 1961 riots and then in the 1964 massacre. Because of the different class structure and race relations in Pemba, there was no local potential for the spontaneous bloodletting there. Revolutionary violence was exported there from Unguja. Regional Commissioners like Rashid Abdalla and his lieutenants unleashed a reign of terror there that is still remembered with horror. A Zanzibar Radio reporter who witnessed this said that not many people were killed in Pemba, but the kind of brutality that was perpetrated there did not occur in Unguja. (Late Abeid, personal communication, Aden, 1966.) The result was a major social dislocation in terms of the number of people killed or exiled, and the persistence of bitterness that continues to poison the current development effort. Estimates of the dead range between 3,000 and 1 1 ,000, according to Clayton. (Clayton 1981:81fn) John Okello, the Ugandan leader of the revolution, states that the total number of those killed was 7,994 a majority of whom were in the rural areas, and they included 1,417 Africans, although the veracity of these rather precise figures cannot be vouched for. In addition 14,716 were detained, 'including women, children, stooges and elderly persons.' (Okello 1967:150-1) Ali Muhsin, leader of the ZNP, on the other hand, has asserted that 13,000 people were killed, 26,000 were imprisoned, and 100,000 were exiled. This would amount to four per cent of the total population killed and 30 per cent exiled, which seem to be on the high side. (Ali Muhsin 1997:35) An American consular official in Zanzibar at the time has tried to compute the approximate figure, and he believes the number of those killed was about 5,000 people, a majority of them being Arab. (Petterson 2002) Those killed, however, were not the large Arab landowners most of whom were absentees living in Zanzibar town, but whole families of poor Manga Arab shopkeepers in the rural areas where they were wiped out, some of them burnt alive in their copra ovens where they tried to hide. Considering the small size of the total population and of the Arab component, anything up to a quarter of the Arab population may have been affected, and it can be fairly described as genocidal in pro-

314

Race and class in the politics of Zanzibar

portions. Whatever the exact number which may never be known, there is no doubting the deep-seated wound that it has left in the body politic of Zanzibar. The primary justification for the revolution was the redressing of imbalances left over by colonialism. Large landholdings were nationalised and distributed to the landless squatters in three-acre plots. Trade was made a state monopoly. Thus, whatever that had remained of the landowning and merchant classes were dealt their final death-blows. Some were killed; many others migrated to the mainland or elsewhere. The Arabs as such ceased to be a credible political force. However, the Revolution was not successful in redressing the imbalance between town and country, and especially between Pemba and Unguja. In fact it exacerbated it. The agricultural sector dominated by cloves, which was monopolised by a state marketing corporation, was squeezed to finance industrialisation and social services, most of which were located in Zanzibar Town (urban housing, TV, etc.) Thus clove producers were paid as little as seven per cent of the international market value of their cloves in 1978, (Salmin Amour 1986:159) and the government still does not want to liberalise the clove trade. A special naval branch of the security forces (KMKM) was set up to chase after clove producers who were trying to smuggle cloves to Kenya to sell at a higher price on the world market than the government's monopoly price. Since 80 per cent of the clove trees were in Pemba, the squeeze was felt there particularly. As a typical rural region, it had been exploited under colonialism as in any other colony, and there was a need to redress this imbalance after independence. However, because a small majority in Pemba had voted for the overthrown coalition parties in the last election before independence, it had to endure decades of exploitation and oppression that is still very vividly remembered. There has been little investment in Pemba which still suffers from poor infrastructure, water and electricity supply, education, etc. Visitors have been shocked by obvious signs of malnutrition among children. The result was a massive migration from Pemba to Unguja and the mainland. Peasants are typically tied to their land, like the proverbial Russian peasants, and this was true of the Pemba before the revolution. The only people from Pemba found in Unguja before the revolution were secondary school students (since there was no secondary school there until the eve of independence) and a few civil servants. The massive uprooting of the people from Pemba is very apparent when we compare the proportion of population in Pemba compared to Unguja. Whereas in 1963 Pemba constituted 48 per cent of the electorate, by 1995 they had declined to a mere 37 per cent. (See Graph 2, S. 316) Many of them migrated to Zanzibar town and the peri-urban area, or to the mainland and beyond. However, by migrating to Unguja in large numbers, they have transformed the socio-political make-up of some of the areas of Unguja that used to be considered safe ASP/CCM seats. During the 2000 elections it is stated that the ruling party was losing even in such strongholds as Raha Leo. By declaring Zanzibar a one-party state and banning the overthrown political parties, the revolution essentially disenfranchised nearly half the population. It was not reconciled to the new order except for a short-lived effort under the second president Aboud Jumbe who attempted a fairer distribution of national development and political expression. Politically, the Pemba continued to be alienated, with the
315

Abdul Sheriff

result that the support for the ruling party declined from 43 per cent in 1963 to a mere 17 per cent in 1995. This massive shift of voter sympathies towards the opposition implied loss of support by the ruling party even among former stalwarts of the ASP in Pemba. Many of the leaders of the opposition party CUF are sons and daughters of the former ASP leaders in Pemba. Diagram 2: Race and Class in Zanzibar

If Zanzibar was split in 1963 by a 54-46 margin, in 1995 the margin was waferthin. Despite all the alleged electoral manipulations, the division between the two parties was almost exactly 50-50. In other words, the Zanzibar polity was divided almost exactly down the middle, even more so than in 1963. That division cannot by any stretch of imagination be considered ethnic or class, and with the massive migration of the people from Pemba to Unguja, the ethnic tendency is becoming less pronounced despite the relative exploitation and oppression of the smaller island.

316

Race and class in the politics of Zanzibar

Bibliography: Ali Muhsin al-Barwani (n.d.), "Conflicts and Harmony in Zanzibar (Memoirs)", Dubai Babu, A. (1991), "The 1964 Revolution: Lumpen or Vanguard?" in A Sheriff & E. Ferguson "Zanzibar Under Colonial Rule", London, James Currey (eds),

Clayton, A. (1981), "The Zanzibar Revolution and Its Aftermath", London, Hurst M. Chilver (eds), "History of East Flint, J. E. (1965), "Zanzibar, 1890-1950", in V. HarlowandE. Africa", vol. II, Oxford, Clarendon Press Lofchie, M. F. (1965), "Zanzibar: Background to Revolution", Princeton, Princeton University Press Mapuri, O. R. (1996), "The 1964 Revolution: Achievements TEMA and Prospects", Dar es Salaam,

Mrina, B. F. and W. T. Mattoke (n.d. c.1980), "Mapambano ya Ukomobozi Zanzibar", Dar es Sa laam, Tanzania Publishing House Okello, J. (1967), "Revolution in Zanzibar", Nairobi, EAPH Petterson, D. (2002), "A Cold War Tale: Americans and the Zanzibar Revolution", forthcoming A. Pirn, (1932), "Report of the Commission appointed by the Secretary of State for the Colonies to consider and report on the Financial Position and Policy of the Zanzibar Government in Rela tion to its Economic Resources", London, Government of Zanzibar Salmin Amour (1986), "The Participation of the Masses and the Role of the Leadership in the Na tional Economic Development, 1964-1984 (A Case Study of Zanzibar)", Ph.D., Karl Marx Party College, GDR Sheriff, A. (1991), "The Peasantry under Impetralism, 1873-1963", in A Sheriff & E. Ferguson (eds), 'Zanzibar Under Colonial Rule1, London, James Currey Tominaga, C. and A. Sheriff (1990), "The Ambiguity of Shirazi Ethnicity in the History & Politics of Zanzibar", Christiantiy and Culture, No.24, (Sendai, Japan), pp. 1-37 Zanzibar Archives, "Thabit Kombo to Ali Sharif Musa", 8.10.1941 , AB 12/2 Zanzibar Protectorate Printer (1924), "Report on the Native Census, 1924", Zanzibar, Government

Zanzibar Protectorate (1931), "Report on the Census Enumeration of the Whole Population, 1931", Zanzibar, Government Printer Zanzibar Protectorate (1948), "Notes on the Census of the Zanzibar Protectorate, 1948", Zanzibar, Government Printer ZEMOG (1995), "Uchaguzi wa Zanzibar 1995", Dar es Salaam, ZEMOG

Abdul Sheriff

Zusammenfassung: Eines der andauernsten Probleme der Geschichtsschreibung Sansibars betrifft die These, dass politische Konflikte aufden Inseln von den Anfangen des Sultanats, uber die Kolonialzeiten bis hin Differenzen entfaltet ha'tten, d. h. in die jungste Vergangenheit, sich entlang von ethnisch-sozialen durch die angeblich strikte ,,rassische" und soziale Kluft zwischen Arabern und Afrikanem. Der Autor analysiert die Wahlen von 1961, 1963 sowie 1995 und zeigt auf, dass das tatsachliche Wahlverhalten der Wahlerschaft nicht mit derartigen Konzepten erklart werden konnen. Er argumentiert, dass nur eine differenziertere, regionale und lokale Unterschiede berucksichtigende Analyse den komplexen Ursachen des aktuellen politischen Konfliktes gerecht werden kann.

Summary: One of the most persistent topics in the historiography of Zanzibar is the argument that from the very establishment of the Sultanate of Zanzibar in the nineteenth century through to colonial times and up to the recent past political conflict in the Islands was essentially informed by ethnic divisions, that is the supposedly rigid Arab-African "racial" and "social" divide. The author examines the 1961, 1963 and 1995 elections, showing that the voting behaviour of the electorate cannot be explained in such terms. He argues that only a much more nuanced analysis, taking into account regional and local divisions, would do justice to the complexities of political conflict and civil strife in Zanzibar.

Resume: Un des themes majeurs de I'historiographie de Zanzibar est la these selon laquelle, depuis les origines du sultanat jusqu'au passe le plus recent, en passant par I'epoque coloniale, les conflits en partant politiques sur ces lies se seraient developpes le long des differences socio-ethniques, du principe qu'un abime pro fond existerait entre les "races" et les "couches sociales" correspondantes. L'auteur examine les elections de 1961, 1963 et 1995 et montre que le comportement de I'electorat ne peut pas etre explique par ces concepts. Seule une analyse prenant en consideration les differences regionales et locales permet de rendre compte des causes complexes du conflit politique actuel.

318