HISTORY OF TANZANIA

State Resilience in Tanzania – Draft Analytical Narrative

I The puzzle: Tanzania’s political stability despite the odds  

Post-colonial Tanzania has remained an enduring “oasis of peace”1 in an East African region characterised by extreme levels of instability and violence.2 The country is comprised of the former Tanganyika and the two Zanzibar islands, together inhabited by 34.6 million people (of which just fewer than 1 million live on the island of Zanzibar).3 Mainland Tanzania was initially a German protectorate (1885-1920) and then became a British-administered territory under a League of Nations mandate (1920-1961). Following the emergence of a strong nationalist movement under the leadership of the Tanganyika African National Union (TANU), its transition to independence was relatively peaceful and untroubled, except for the army mutiny of 1964 that was suppressed with the help of the British. The four decades since independence have been characterised by remarkable levels of peace and stability. While the socialist policies in the wake of the “Arusha declaration” (1967) involved the use of considerable state force (most notably against the owners of cash-crop producing land), largescale violence remained absent. Significantly, political stability was also maintained during the severe economic crisis of the 1980s, some cases of violence against “economic saboteurs” and a failed coup d’Etat being rather minor exceptions. While subsequent economic and political liberalisation has been accompanied by increasing political and socioeconomic tensions, they have largely remained non-violent. Despite the “violence of everyday life”4 and the what some might call the “structural violence” 5 of enduring poverty, mainland Tanzania remains light-years away from the violent excesses experienced by many of its neighbouring countries and continues to provide a significant degree of security to its citizens – the “social protection” that Karl Polanyi said is so crucial to human aspirations everywhere.6
Altogether, Tanzania is an amazing case not only of state survival, in a sense, against the odds, but also of a society and polity, which have enjoyed relative peace and stability for more than four decades since independence. This remarkable stability should not – and was not – taken for granted. When, in the mid-1970s, ten years after independence, President Nyerere was asked by a journalist, “what was your greatest success?”, he replied, “That we have survived”.                                                  1 Hofmeier 1997, 151 2 This is especially true for Mainland Tanzania, to which we explicitly limit our analysis. On the Zanzibar islands (Unguja and Pemba), by contrast, political stability has been more volatile. Both the pre-colonial Oman Sultanate and the British Protectorate (1890-1963) were characterised by severe racial tensions between the ruling Arab/British minority and the black majority. Only one month after independence (1963) these tensions culminated in the Zanzibar Revolution, which established the leadership of the Afro-Shirazi Party (ASP) and caused several thousand (mainly Arab) deaths (Hofmeier 1997, 153). After Tanganyika and Zanzibar united (1964) and the ASP and TANU merged to form the Chama Cha Mapinduzi (CCM) (1977), the situation improved gradually throughout the 1970s and 1980s. With the introduction of multi-party politics (1992), however, the old tensions started to resurface and culminated in violent clashes around the 1995 and 2000 elections (Kaya 2004, 126ff.). Despite this recurrent and largely unresolved conflict, even Zanzibar remains far from the large-scale violence observed in many other (East) African countries (e.g. Rwanda, Burundi, Uganda, DRC). 3 EIU 2007, 16 4 On the violence of everyday life see Heald 2002.  5 Some might argue that the extreme poverty experienced by Tanzanian people is a form of “structural violence”, in the manner in which Peter Uvin (1998) discussed this in relation to Rwanda. However, we would argue against conflating the effects of poverty with the effects of the physical violence that accompanies civil war or the free reign of local warlords or armed gangs. 6 Polanyi 1944, 26. See also Putzel 2002.

When asked the same question by the same journalist in 1983, he said, “I think I would still give the same answer”.7 The puzzle that therefore requires explanation in Tanzania is how the state survived and presided over a generally peaceful management of conflict despite not only profound economic crisis, but also surrounded by states experiencing insurgency and civil war.
Our main proposition is that Tanzania’s political stability can be explained by the ruling party’s ability to forge and maintain a centralised and inclusive political coalition. This form of political organisation can theoretically be termed an “inclusive elite bargain”.8
Tanzania can be characterised as part of a large set “neopatrimonial states” but what is specific about the organisation of patronage in Tanzania is that it is “centralised” as opposed to framed within our literature review. It is inclusive of all significant sources of political, military, economic and social power within the territory of the Tanzanian state. This type of centralised and inclusive bargain is what accounts for state resilience (also termed “political stability” also termed “peaceful management of conflict”). This form of political organisation has been able to establish the hegemony of state institutions (rules) over and above other potentially rivalling institutions whether emanating from traditional, religious, regional or ethnic institutional frameworks. In this sense the Tanzanian state has significantly reduced the negative consequences of “institutional multiplicity”. We suggest that this form of political organisation is responsible for the state’s ability to endure repeated episodes of economic crisis, persistent poverty, destabilising pressures of structural adjustment and potentially destabilising security challenges (wars in, and refugee flows from, neighbouring countries). By focusing on the elite bargain, we do not by any means imply that what is happening in society as a whole, or in non-elite circles, is unimportant. Elites by definition are part of society as a whole and they need to maintain their legitimacy through coercion and persuasion within society. They are only “elites” because they command authority in society so our exploration of evolution of the elite bargain involves necessarily an analysis of social evolution over time.
We test this proposition by examining the process of state formation in Tanzania over time and the manner in which the state managed economic, socio-political and security crises during each historical period.  

2 Pre-colonial and colonial legacy  

The parameters of the post-independence Tanzanian state were clearly drawn during the relatively short period of colonial history, roughly 30 years of German rule followed by forty years under British authority. While the legacy of colonial rule was relatively “favourable” in that segregation between indigenous groups remained low, the nationalist movement skilfully built upon this situation by rallying the main social forces in Tanganyikan society and thereby laying the basis for the “inclusive elite bargain”.
The pre-colonial legacy
Political arrangements in the territory of pre-colonial Tanganyika varied widely related to the patterns of geography, production and trade, where forms of organisation included anything
                                                 7 Cited by Shivji (1990), referring to Africa Now, vol.32 (December 1983), p.101.  8 Cf. Putzel 2007, DiJohn 2008, Lindemann 2008.

from the acephalous to highly structured hierarchical chiefdoms.9 While most political systems were composed of small, autonomous communities of crop cultivators or pastoralists organised at the clan or kin-group level, there were also several more centralised forms of political organisation with strong rulers presiding over larger communities (e.g. the Chagga, Hehe, Ngoni, Gweno, Nayamwezi, or Sangu).10 On the whole, most pre-colonial societies were not based on clearly defined tribes but rather characterised by enduring processes of inward migration and colonisation whereby social and language groups “merged imperceptibly” into one another. 11 Ethnic consciousness was therefore relatively un(der)developed and chiefs were mostly not very dominant or powerful.
The territory that would become known through European colonial rule as Tanganyika contains a total of about 120 of these fluid ethnic groups (cf. map 1). However, no one group has a dominating position in the population (e.g. as the Buganda in Uganda). The largest ethnic groups include the Sukuma (13%), Makonde (4%), Chagga (3,7%), Haya (3,5%) and Nyamwezi (3,4%).12 This diverse, yet no segregated nature of Tanzania’s ethnic composition has arguably facilitated – yet by no means predetermined13 – the achievement of unity during both the anti-colonial movement the post-independence administration (see below).
German colonial rule (1985-1920)
As German interests began to establish their authority over Tanganyika from 1895, they created a colonial order that would severely weaken the political, economic and military structures of pre-colonial societies. At the same time, and this is key to our argument, the new order did not entail significant imbalances between the multiple ethnic groups. While the colonial administration heavily discriminated against Africans in favour of European interests and Asian traders, it did not create systematic socio-economic segregation among African groups.14
Politically, the territory was divided into 22 administrative districts with considerable power left in the hands of district officers backed up with military force.15 Even though the colonial state never had full control over the territory of Tanganyika, the power of traditonal authorities was further weakened by a mixture of “semi-indirect” and (mostly) “direct” rule.16 Under “semi-indirect” rule, the Germans made use of chiefs but stripped off much of their traditional powers, even changing their traditional names. “Direct” rule, by contrast, was applied where there were no strong traditional rulers. Initially, the colonial administration simply took over the Sultan of Zanzibar’s officials: liwalis in the main towns, akidas under them, and jumbes responsible for tax collection at the local level.17 Later on, the Germans replaced the liwalis with German officials and appointed younger African graduates as akidas                                                  9 Iliffe (1979, 21) is careful to point out that what he calls “stateless” communities were generally not “less advanced” than the chiefdoms – many of the “most cultured peoples were stateless”. 10 Mpangala 1999: 9. With the exception of the small kingdoms of Buhaya, Tanzanian societies had therefore not developed highly centralised states like those of the Interlacustrine Region (e.g. the Buganda or Bunyoro). 11 Iliffe 1979, 8f.  12 Hofmeier 1993, 180 13 Some scholars have argued that high degrees of ethnic fractionalisation necessitate coalition-building and therefore constitute – by themselves – a source of safety for the most heterogeneous countries (Elbadawi & Sambanis 2000; Collier & Hoeffler 2002). However, we argue that ethnic cleavages are always contingent on inclusive vs. exclusionary political organisation. This is why the trajectories of highly heterogeneous countries range from enduring political stability (e.g. Tanzania) to recurrent civil war (e.g. Chad). 14 Klugman et al. 1999, 77f. 15 Iliffe 1979, 118f.  16 Mpangala 1999, 13 ff. 17 Coulson 1982, 41f.

and jumbes. Significantly, African civil servants at the middle- and lower level were recruited from various ethnic groups. Also, since the Germans used the Sultan’s bureaucracy, the also used his language in district administration, Swahili – a decision that would later be endorsed by the British18. They thus introduced a national language that would facilitate inter-ethnic dialogue and greatly contribute to the formation of a Tanzanian national identity.19

Map 1: Tanzania’s ethnic groups

Economically, German colonisation – combined with the devastating impact of several epidemic diseases – transformed earlier patterns (where Zanzibar interests had controlled trade) effectively putting “an end to the prosperity of indigenous colonial economies”.21 While investment in the manufacturing sector remained limited, the Germans opted for a mixture of settler and plantation agriculture on the one hand, and a cash-crop oriented peasant                                                  18 Iliffe 1979, 208ff., 529; Coulson 1982, 42. 19 Swahili does not belong to any particular ethnic group that could use it for dominance (Omari 1995, 28). 20 Martin 1988: 21. 21 Hyden 1980, 41. Cf. also Iliffe 1979, chapter 5.

agriculture on the other. Although land alienation never reached the same extent as in Kenya, colonial policy generally sought to establish a settler and plantation agriculture where African labour was used to extract revenue from cash crops (sisal, rubber, cotton, coffee).22 Beyond these plantation and settler enclaves, the colonialists promoted a “pre-capitalist peasant mode of production”, in which the incorporation of rural producers into the wider cash-crop oriented national economy was an essential element.23 Taxation measures (essentially a hut tax) were introduced in 1897 only to force people into the money economy, delivering surpluses to market and seeking wage employment.24 While plantations and export-import companies were mostly owned by the Germans or other Europeans, Asian traders were actively encouraged to capture large parts of domestic commerce.25
The structure of the colonial economy had two important consequences. First, it contributed to the weakening of traditional authorities, not least of those who had been relatively strong in pre-colonial times. The fact that Germans made use of stronger chiefs in tax collection (“semi-indirect” rule as outlined above) not only meant a loss of income for the chiefs (who had received tribute before) but also undermined their own standing in their communities.26 Second, and probably more importantly, the lack of alternatives to independent peasant production or wage labour on plantations strengthened the socio-economic homogeneity of the African population.27 While regional imbalances therefore remained far less pronounced than in other colonies (e.g. in Uganda), colonial rule nonetheless lad the foundation for a certain degree of uneven development. Infrastructure was mainly develop to support exportcrop production along the Central Line (sisal) and in the northern areas (cotton, coffee), whereas the other regions were treated as areas of labour supply.28
Finally, German colonial rule also significantly undermined the military capacities of indigenous societies. The armed movement that left the greatest legacy in this respect was the Maji Maji rebellion that lasted from 1905 to 1907 – the last major effort of pre-colonial authorities to resist the colonisation process. The rebellion was rooted in stateless societies of the south and self-consciously opposed the people’s incorporation into the colonial economy (in particular the compulsory cotton scheme). It was brutally suppressed by the German authorities (with about 75000 deaths) and was followed by three years famine.29 The Maji Maji revolt had two lasting effects. First, it transformed the political landscape of the south and virtually eliminated Ngoni military society,30 which might have served as a base for future armed opposition to colonial authority. But, even more importantly, the devastating impact of the rebellion created lasting memories among traditional leaders who were loath to turn to armed resistance later during the British colonial period. In a sense, the scale of defeat, underpinned thinking about non-violent routes to independence in the 1950s.
British colonial rule (1920-1961)
                                                 22 In 1913, there were 5336 Europeans in the country, more than in Kenya at that time (Coulson 1982, 40). The settler population had achieved significant power in the colony and were arguably well on the way to making it another Rhodesia. 23 Hyden, 1980, 42f. The integration into the capitalist economy was however marginal in that the pre-colonial modes of production were replaced by a new pre-capitalist mode based on independent peasant production (combing both cash-crop and subsistence oriented agriculture). 24 Iliffe 1979, 133. 25 Coulson 1982, 39f. Cf. also Iliffe 1979, chapter 5. 26 Iliffe 1979, 133; Coulson 1982, 35. 27 Klugman et al. 1999, 77. 28 Harnevik et al. 1988, 12. 29 Coulson 1982, 31. 30 Iliffe 1979, chapter 6.

The British take-over after the First World War reinforced the existing political and economic structures in Tanganyika. Despite introduction of “indirect rule”, traditional authorities remained weak, while British economic policy continued to prevent – at least until the mid1950s – the emergence of major socio-economic segregation among Africans.
At the political level, the onset of British rule was soon followed by a shift to “indirect rule”. The former Governor of Nigeria, Sir Donald Cameron, was appointed to head the colonial government in Tanganyika in 1924, whereby the British opted to adopt outright the system they had established in Nigeria. Indirect rule meant the integration of “indigenous political systems” into the colonial administration through the establishment of a “native authority”, which was some combination of a chief and a council, native courts and a native treasury.31 As a consequence “Tanganyika experienced a vast social reorganisation in which Europeans and Africans combined to create a new political order based on mythical history”.32 Where German colonial rule had exercised direct control, the British “invented” traditional chiefs and gave them indirect rule powers and authority.33 In situations where Germany had relied on “semi-indirect” rule, by contrast, the British provided the traditional chiefs with more power and restored their traditional names of chiefly authority. After the Second World War, an attempt was made to further develop “indirect rule” by introducing a system where smaller chiefdoms were placed under “paramount chiefs”.34 Altogether, the invention of new tribal identities was the result of both British officials and Africans wanting more effective units of action and was widely perceived as a route to stability.
While the introduction of “indirect rule” seems to imply a revitalisation of indigenous political authorities, the available evidence indicates that the system of “ruling through the chiefs” remained largely ineffective.35 First, the search by the British administrators for the “legitimate” chief was often fruitless. While northern Nigeria or in Buganda exhibited the existence of highly centralised and powerful chiefdoms, in Tanganyika the same degree of centralisation had either never been achieved or been weakened by German rule (cf. above). Second, the amount of real power held by the chiefs was never great as the Provincial Commissioners and District Commissioners were always at their side. In this sense, indirect sense “was a transparent attempt to disguise the reality of foreign rule”.36 Third, and certainly not least, indirect rule was thwarted by the development of a colonial economy that reduced the real power of chiefs (see below).
Economically, Tanganyika was “at the bottom of the imperial pecking order”. 37 Its protectorate status implied a time-constrained period of colonial rule and this meant that a significant transformation of the country’s economic systems was even less a priority for the colonial state compared to other colonies.38 Even though the British would eventually                                                  31 Iliffe 1979, 319f. This stood in contrast to both the former German system, where native authorities appointed underneath district officers needed to have no status in their communities and with the model of a “native state” like that which existed in Buganda, “where relations between European and African rulers were fixed by treaty”. 32 Iliffe 1979, 324. 33 Mpangala 1999, 15. 34 Mpangala 1999, 17f. The Chagga, for instance, originally constituted a number of smaller chiefdoms around the Kilimanjaro Mountain and were now united under one Chief or Mangi, the Mangi Mareale. 35 Coulson 1982, 96 36 Coulson 1982, 97 37 Iliffe 1979, 302 38 “During the 1930s Britain was a net importer of capital and migrants. Of course, even Britain’s waning economic vigour could have transformed Tanganyika had the territory possessed any economic attraction, but it offered nothing to compare with South African gold, Middle Eastern oil, or Malayan tin and rubber. Continuing poverty was British Tanganyika’s leading characteristic.” (Iliffe 1979, 261)

implement rudiments of an import substitution oriented industrial polity after the Second World War, there was initially a conscious decision “not to industrialise”39 and investments in infrastructure remained limited.40 With respect to agriculture, the British generally maintained the balance between peasants, settlers, and plantations inherited from the Germans.41 While small-holder agriculture was consolidated and received little challenge from large-scale settler agriculture,42 the introduction of new commercial crops (coffee, tobacco) led to a limited process of agricultural diversification. 43 From the mid-1930s, the colonial government considered that an increasing use of force was necessary to make small peasants change their agricultural techniques.44 It therefore used the Native Authorities to implement compulsory agricultural development schemes for the cultivation of export commodities and to impose monopoly marketing arrangements. While the increasing commercialisation of agriculture gave rise to “cooperatives” representing the emerging African capitalist agriculture (the “Kilimanjaro Native Planters’ Association” was founded as the first cooperative in 1925), the latter were for at long time contained by a coalition of settlers, chiefs and the administration.45 It was only in the early 1950 that the British finally began to promote African commercial farming by allowing African cooperatives to develop and concentrating resources on large farmers (“kulaks”) – a move that was largely motivated by the desired to strengthen social control and bolster the colonial mode of production.46
Economic developments under British colonial rule had two important consequences. First, the implementation of coercive agricultural development schemes through the Native Authorities again discredited the chiefs and unmasked them as the stooges of the British.47 As agricultural policies grew increasingly unpopular, the position of the traditional authorities deteriorated. Moreover, the commercialisation of agriculture led to the development of new forms of organisation among Africans eventually further marginalizing traditional authorities. While chiefs were initially were initially able to contain or control African cooperatives with the help the colonialists, this changed from the 1950s when the new elite used the cooperative to take the leadership from the traditional aristocrats. Second, British policies ensured that socio-economic segregation among the Africans developed only towards the end of colonial rule when the rise of the cooperatives “led to regional imbalances and increased inequality among Africans”.48 But even though capitalist African agriculture had taken root, this group, consisting only of a few hundred farmers had very little political clout.49
In the security realm, finally, the British did not recruit along ethnic lines, pitting those who collaborated against those who resisted their authority, as was the common practice in many other colonies. Omari speculates that this was because the Germans had already pacified the
                                                 39 Coulson 1982, 71. The British feared that industry would bring ruin to the tribal society that “indirect rule” was trying to retain - “detribalised” African workers in the towns were perceived as “dangerous”. 40 The British period saw very little investment in the kind of infrastructure necessary for development. Between 1938 and 1960, road mileage increased from 15,000 to 20,464, (only 27%) with main roads being extended from 2,700 miles to only 3,770. The construction of railways was even less significant with mileage increasing from 1,370 to only 1,760 miles (only 22%) (Stephens XXX, 82). To be added to the bibliography. 41 Iliffe 1979, 201ff., Coulson 1982, 45 42 Settler farms received limited state support and expropriation of land remained relatively low, especially when compared to neighbouring Kenya. 43 Iliffe 1979, 286ff. 44 Coulson 1982, 52 45 Coulson 1982, 60ff. 46 Coulson 1982, 55ff. 47 Coulson 1982, 52 48 Klugman et al. 1999, 79.  49 Khan & Gray 2006, 36

colony and there was no need for the British authorities to do so.50 While the armed forces created by Britain were ethnically unbalanced, with members drawn from only seven ethnic groups (5% of all ethnicities of the time), there was no one ethnic or regional group dominating the military. The British stationed units of the King’s African Rifles in Tanganyika under the command of British officers, just as in Uganda.
The nationalist movement
Colonial rule in Tanganyika left a relatively “favourable” legacy in that fragmentation between indigenous groups remained low, not least when compared with neighbouring countries (DRC, Uganda, and Rwanda). Nonetheless, one can hardly overemphasise the achievements of the nationalist movement that skilfully rallied the main social forces in Tanganyikan society and thereby achieved a comparatively high degree of unity upon independence.
The nationalist movement in Tanganyika had three main pillars.51 The first pillar was the Tanganyika African Association (TAA) established in 1929 to rally all Africans against colonial domination. It was initially an urban-based movement of civil servants and teachers that had grown out of the Tanganyika Territory African Civil Service Association (1922). While the TAA was in a rather weak position in the late 1930s, it was revitalised by the entry of young intellectuals in the early 1950s, most notably Julius Nyerere who was elected president in 1953. From then on, the TAA took the lead in the anti-colonial struggle, especially after being transformed into a nationalist political party in 1954 – the Tanganyika African National Union (TANU). The second pillar was the trade-union movement that originally developed in opposition to the allegedly elitist stance of the TAA, led by a former leader of the African Welfare and Commercial Association. Union-based protest against British rule increased throughout the colonial period and manifested itself in the dock strikes of 1937-39, the general strike of 1947 and the major riot of 1950. While the latter incident led to the temporary ban of the union movement, it was allowed to reorganise under the Tanganyika Federation of Labour (TFL) in 1955. Union membership increased from 13.000 in 1956 to about 200.000 in 1961.52 The third pillar was the cooperative movement that gained momentum in the rural areas from the early 1950s (see above) and demonstrated African ability in business.53 Between 1952 and 1952, the number of cooperatives rose from 188 to 617, while their membership increased from 153.000 to 325.000.
From the late 1940s and early 1950s, the TAA started to assume an integrating role in the nationalist struggle. As it began to develop ideas of struggling towards national independence, it had to transform its character of an urban elite movement and find ways to sink its roots more deeply into Tanganyikan society. This was done by simultaneously tapping into movements for tribal reorganisation and bringing together the incipient nationalist movement found in cooperatives and trade unions. Tribal reorganisation began during the 1940s and 1950s as struggles of a number of ethnic groups against the cash-crop oriented colonial policies.54 While these “proto-nationalist organizations” could have developed into ethnic political parties (as in other colonies), the TAA managed to co-opt them into the nationalist                                                  50 Omari 2002, 92f. 51 Coulson 1982, 102 ff.; Klugman et al. 1999, 79f.  52 Harnevik 1993, 32. By 1961, 42% of the country’s workers were unionised compared with 12% in Uganda and only 8% in Kenya.  53 Cooperatives were especially popular among the African population because they ended the long-resented Asian monopoly on the trading circuits. 54 Iliffe 1979, 495 f.; Mpangala 1999, 19 f. Examples include the Sukuma Union formed in 1945, the Pare Union in 1946, the Chagga Kilimanjaro Union in the late 1940s, and the Meru Union in 1951.

movement (whereby they became branches of TAA). As tribal improvement societies failed to make real gains for their communities, the TAA won young tribal activists to the nationalist movement, often at the expense of the already discredited traditional authorities (see above).55 This process incrementally expanded the legitimacy of the nationalist movement to a wide cross section of the population. Nyerere’s own origin as a son of a chief from a small ethnic group, the Zanaki, played a crucial role in this context. When Nyerere advocated unity, other ethnic leaders did not see him as an immediate threat to their own position.56
The role of the TAA was further strengthened when it was re-launched as TANU in 1954, inspired by the models of the British Labour Party and the Ghanaian Convention People’s Party. This reinvented organisation had three key attributes:57 (1) the organisation was explicitly committed to the goal of independence and proclaimed its opposition to tribalism as well as its commitment to elections, education, trade unions and cooperatives and wider goals of African liberation; (2) unions and associations (whether based on tribe or occupation) could join TANU and pay a “political levy”, but TANU had its own branches of youth and women’s organisations, which Nyerere saw as central to the organisation’s work; and (3) an annual delegates conference would elect the national executive committee, giving significant representation to the base, but day to day activities would be carried out by elected officers and a central committee appointed by the TANU president, which allowed Nyerere the space to co-opt in notables and build wider unity.
Altogether, TANU managed to forge a highly inclusive coalition of elites that transcended ethnic, racial, religious and occupational boundaries. First, it implemented its outright multiethnic approach and skilfully incorporated key figures from various ethnic groups into the movement.58 Second, the party rejected any form of racialism and actively welcomed whites and people Asian origin. Third, and in contrast to division in Uganda, the independence movement in Tanganyika was built on religious tolerance. It was “led by Julius Nyerere, a Catholic, Rashidi Kawawa, a Muslim, and Oscar Kambona, an Anglican”.59 Fourth, TANU also managed to secure the support from leaders of trade and cooperative unions, even though there was no open cooperation until 1957 (as unions and cooperatives were watched over by the colonial government).60 Beyond this inclusive coalition of elites, TANU also developed into a powerful mass-based political party that rallied thousands of peasants, workers and traders.61 Between 1954 and 1960, TANU membership grew from 15.000 (mostly transferred from TAA) to about 1.000.000, or one in five adults.62
The colonial administration made an attempt to divide the nationalist movement by supporting the creation of a rival party in 1956, the United Tanganyika Party (UTP),                                                  55 Often the TAA was able to bring national and even international attention to local struggles over land rights. 56 Omari 1995, 27. Nyerere was outstanding in using every opportunity to incorporate factions into the nationalist movement. It is doubtful whether without Nyerere’s skill and creativity, TAA/ TANU would have remained so relatively free of internal disunity. 57 Iliffe 1979, 507ff.. 58 Omari 1995, 26. Key ethnic leaders that joined TANU included Chief Fundikira of the Wanyamwezi, Chief Anna Gwassa of Kasulu, Chief Therea Ntare of Heru Kasulu, and Chief Kasusura of Rusubi-Biharamulo. 59 Kaya 2004, 164. Christian and Muslim proselytisation had continued apace throughout the British period, when by the late 1950s, Muslims constituted some 31% of the population, Christians were up from 2% in 1914 to 10% in 1938 and 25% by 1957 (17% Catholic and 8% Protestant). Due to attitudes of the colonial authorities, Muslims were more uniformly pro-TANU and Christians more divided between support for the colonial government, often allied to traditional authorities, and support for TANU. 60 Coulson 1982, 107 61 The role of traders was particularly important as they carried TANU’s messages to distant communities in Swahili and they were able to link up local grievances with nationalist demands for independence. 62 Coulson 1982, 115; Iliffe 1979, 536.

essentially composed of disgruntled chiefs and white settlers.63 Also, a split from the TANU leadership led to the creation of the African National Congress (ANC) that represented Africans only and thereby challenged TANU’s ideology of unity and equality.64 Both parties were however severely defeated in the elections of 1958 and 1960 when TANU secured sweeping victories. While TANU therefore led the country to independence in a position of political power, it had not developed a military orientation, partly as a result of earlier defeats and partly because of the influence of Ghandi’s philosophy of non-violent resistance.65 Security matters can therefore be regarded as TANU’s “weakest link” – a weakness was soon to be confirmed by the post-independence mutiny of the armed forces (see below).
With independence and the expansion of the franchise in many former colonies in Africa, democratic forms of government opened the way for factionalism and the politics of regional ethnic and language groups. John Iliffe concluded that there were three reasons why this did not have the disintegrating influence in Tanzania that it had elsewhere. First there was no ethnic group large enough in Tanzania to think seriously about separation. Secondly, the Swahili language bridged the divide between ethnic groups. But thirdly, and what seems most importantly, TANU had built a strong party machine throughout the territory before the franchise was expanded.66 When TANU faced a much expanded electorate after the 1962 Constitution was adopted, it won an even greater margin of victory than in 1958 under a limited franchise.  

III Independence to Arusha Declaration (1961-1967)  

The unity of the nationalist movement notwithstanding, it soon became clear that the country’s political stability remained fragile. As everywhere across the continent, independence created enormous expectations among peasants, workers and civil servants who longed for social services, employment opportunities and higher incomes. Failure by the postcolonial government to immediately meet these demands was bound to create disunity and unrest. Accordingly, the early 1960s saw mounting confrontations with radicalised trade union leaders who mobilised as series of strikes for wage increases and rapid Africanisation. More importantly, the fragility of political stability was underlined by the violent overthrow of the government of Zanzibar in early 1964, almost immediately followed by the mutiny of the Tanganyikan armed forces that shook the country in its foundations and was suppressed only with the help of British troops.67 Interpretations on how much of a threat the military mutiny represented to the TANU-led government vary considerably. Bienen emphasises that the mutineers never came close to seizing power.68 According to him, the rebels repeatedly emphasised that they did not want a coup, but only Africanisation and improved wages. The Minister of Defence, Oscar Kambona, was instrumental in negotiating with the mutineers and
                                                 63 Klugman et al. 1999, 80; Coulson 1982, 115 64 The ANC had been founded by an Assistant Secretary General of TANU, Zuberi Mtemru, which broke from TANU over participation in the racially segregated colonial election of 1958 (Baregu 1994, XXX). 65 Lupogo 2001. 66 Iliffe 1979, 570. 67 On 19 January 1964, led by a non-commissioned officer, the first battalion left its barracks and took key installations in Dar es Salaam: radio station, police stations, the airport and State House. They arrested British officers from their homes and barracks. Initially, the mutineers demanded the expulsion of British officers and higher pay in negotiations with TANU ministers (both the President and Vice-President went into hiding). The second battalion, based in Tabora, joined the mutiny two days later. In Dar es Salaam, looting broke out almost immediately after the mutiny began, especially as the police decided not to intervene. Some soldiers were involved in the looting and mainly Asians and Europeans were targeted. 68 Bienen 1978, chapter 7. To be added to the bibliography.

the President treated the mutiny almost like a labour dispute, consciously playing down its implications. TANU ranks remained united and the civilian posts it controlled continued to function throughout. Civil servants did not join in, nor was the mutiny endorsed by key traditional leaders. It was the strength of the TANU organisation that prevented the mutiny from turning into a full blown coup d’etat. Bienen’s emphasis on TANU’s strength is at least questionable. Swai, for instance, claims that the mutiny underlined weakness and fragility of the political system.69 It was organised by a “handful local rank and file who were poorly trained and armed and yet managed to force the whole government machinery to a standstill for several days”. Despite its organisational strength, TANU was unable to organise any local resistance and was rescued by “only a handful of British marines”.
Not matter how severe the threat posed by the mutiny, it is clear that Tanganyika’s first years of independence were altogether relatively peaceful and untroubled, especially when compared with the violent excesses in neighbouring Congo or Rwanda. This relative political stability between 1961 and 1967 can in part be explained by the pattern of colonial rule and the achievements of the nationalist movement. More important, however, was the fact that TANU and its leader, Joseph Nyerere, took a series of decisive steps to consolidate the “inclusive elite bargain”. TANU was further developed into a centralised and inclusive political organisation that imposed its hegemony over Tanganyikan society and introduced beginnings of rent-sharing policies. Perhaps even more importantly, it used the occasion of the army mutiny to completely reorganise the security sector. The combination of these actions proved crucial to the state building effort and would have a lasting impact on the shape of Tanzanian state.
POLITICS
One, if not the major achievement during the first years of independence was TANU’s consolidation into a powerful ruling party that would become coterminous with the Tanzanian nation state. Significantly, the party’s dominant position hinged on a skilful combination of both centralised authority and inclusive cooptation. In the words of Dashwood & Pratt, it developed an “original set of political institutions” that were “sufficiently participatory and democratic as to limit the risk of the regime’s becoming severely authoritarian and corrupt, but not to open as to threaten Tanzania’s still fragile unity or weaken prematurely the integrating and energising capabilities of the nationalist movement”.70
TANU’s centralised authority was first established at the expense of rival institutions thereby further minimising “institutional multiplicity”. Most importantly, “two fundamental institutions of indirect rule” were abolished in 1962.71 While the Native Authorities were disbanded and replaced by elected District Councils, the chiefs – already weakened during the pre-independence period – now lost all official power that was transferred to politically appointed Regional and Area Commissioners.72 At the same time, their status was further undermined by the government’s decision to nationalise all land (see below). Even though TANU faced resistance in some regions where its party organisation was still weak, most                                                  69 Swai 1991, 95. 70 The ruling party also maintained a hard line towards other organisations that would not submit to its authority. Religious organisations were tolerated if they did not challenge the political status quo. However when they did, they were banned, as happened to the East African Muslim Welfare Society in the late 1960s (Kaya 2004, 150). While some argued this tight control was evidence of authoritarian rule, it could as well be interpreted as a key measure that avoided the radicalisation of Islam in Tanzania (Chris Peter, interview). 71 Coulson 1982, 136 72 While tribal authorities were removed, customary law was maintained though adjudicated through the normal courts (Chris Peter, interview).

traditional authorities tried to ensure their position by joining the organisation. But this led to serious conflict at times. In one district commoners were elected as TANU representatives and one of their first actions was to reduce onerous levels of bridewealth and ban the use of cattle in paying bridewealth or as collateral for loans. This challenged the position of tribal elders and brought the district to the “fringe of civil tribal fighting”.73
Furthermore, TANU sought to assert its hegemony vis-à-vis the trade unions and the cooperative movement – the other two main pillars of the nationalist movement (see above).74 With 200.000 members by 1961, the trade unions had grown into a powerful social force.75 While moderate union leaders had been co-opted as cabinet ministers, the new (more radical) leaders mobilised as series of strikes for wage increases and rapid Africanisation. As a consequence, four acts were passed in 1962 that limited the right to strict, introduced prevention detention76, prevented civil servants from joining unions and gave TFL greater power over its constituent unions. This conflict culminated into the 1964 decision to replace with the National Union of Tanganyika (NUTA) that was subordinated to TANU and involved compulsory membership for workers. 77 Gaining control over the cooperative movement, by contrast, proved to be more difficult. The number of cooperatives increased from 857 in 1961 to 1533 in 1966.78 While the government controlled their registration and intervened in their operations in many ways, the movement managed to preserve a certain degree of autonomy throughout the 1960s79 – a situation that motivated TANU’s efforts for further control during the 1970s (see below).
TANU’s power was also asserted over other organs of the state. The ruling party was centralised and highly structured with party organs at all levels of society from village cells of 10 households up to the National Executive Committee (NEC). The “party structure itself was transformed into a bureaucracy, and integrated with other bureaucracies such as the hierarchy of regional and district commissioners” thereby constituting a prime example of what Allen describes as “centralised-bureaucratic politics”.80 Given this unusual degree of organisational cohesiveness, TANU by far overshadowed other state institutions. The National Assembly, for instance, was clearly subordinate to the ruling party: Decisions were typically prepared and taken by TANU and then “rubber-stamped” in Parliament.81 In 1965, Tanzania was formally transformed into a one-party state, thereby further consolidating TANU’s now virtually uncontested position on the political playing field. From now on, all political competition would take place within the realms of the ruling party.82
But TANU’s dominant political position did not simply rest on centralised authority. Instead, it also involved a significant degree of inclusiveness. To begin with, the one-party state was “democratic” in that it allowed for inner-party electoral competition. According to the 1965 constitution, (1) TANU membership was open to anyone willing to accept its objectives; (2) any TANU member could be nominated for election to the national assembly and party
                                                 73 Iliffe 1979, 570. 74 Dashwood & Pratt 1999: 249f. 75 Coulson 1982, 137ff. 76 The “Preventive Detention Act” (1962) was repeatedly used against trade-unionists and underscored TANU’s “authoritarian face”. Christopher Tumbo, the leader of the railway union, was detained without trial for four years (Coulson 1982, 140). 77 Bienen 1967, 271f. To be added to the bibliography. 78 Coulson 1982, 149 79 Harnevik et al. 1988, 57. 80 Allen 1995, 305f 81 Harnevik et al. 1988, 95 82 Kaya 2004, 152.

organs; (3) in each constituency, the annual district conference selected two nominees among the candidates (that were screened but normally accepted by the NEC); and (4) local candidates were not allowed to spend money on campaigning or to use tribal, racial or religious language.83
Moreover, TANU built on its achievements during the nationalist struggle and maintained an inclusive elite coalition. The first Nyerere Cabinet included intellectuals, rural activists from the cooperative movement, urban labour, chiefs 84 , “accommodating immigrants” (i.e., Europeans) and TANU professionals. Also, having to pursue an Africanisation programme upon independence since representation of Africans in the civil service or among university students was appallingly low,85 TANU resisted any form of ethnic or racial favouritism.86 As a consequence, the Tanzanian elites remained largely representative of ethnic, regional and religious divisions.87 With respect to ethnicity, 37% of all ethnic groups were represented in the leadership, while neither one single nor any of the large groups enjoyed serious overrepresentation. As for regional representativeness, every region was represented in the leadership, even though the northern regions (Kilimanjaro, Tanga, and West Lake) as well as Zanzibar-Pemba benefited from overrepresentation. Religious divisions, by contrast, were less equally reflected in the Tanzanian leadership. While traditional religions were not represented, Christians enjoyed overrepresentation relative to the Muslims.
Beyond this balanced distribution of leadership positions, TANU adopted a series of further measures to rule out the exploitation of ethnic, religious, or regional identities. Not only did the government ensure that civil servants would serve in districts other than the one they hailed from, but it took the radical policy of ensuring that at least half of those attending secondary school would do so in a different district than the one in which they grew up.88 Also, the government greatly expanded policies begun under the German and British authorities to promote Swahili as the language of instruction among the different ethnic groups found in Tanzania and Swahili was also used as the language of government.89
ECONOMY
Upon independence, the Tanganyikan economy was largely agricultural, with the sector representing about 45% of reported output and manufacturing industry no more than 7%.90
                                                 83 Dashwood & Pratt 1999: 241f. Cf. also Pratt 1976, 201ff.  84 While TANU had stripped the chiefs of all official power, it nonetheless continued to work hard to ensure that notables had a major place in its organisation. Most notably, Chief Fundikira had a role in the TANU leadership and spoke on behalf of traditional authorities. Also, it is important to keep in mind that the nationalists were often themselves descendents of traditional authorities. When Nyerere argued for the elimination of tribal authority, he did so as the son of a chief and his actions meant his own brother would not assume the leadership role that would have been coming to him (Chris Peter, interview). 85 At independence no African held the position of provincial commissioner; only 2 of 57 district commissioners were Africans; Africans had only 1,170 of 4,452 senior civil service posts. In the professions, only 16 of the 154 doctors in the territory were Africans; 1 of 84 civil engineers and only 2 of 57 lawyers (Iliffe 1979, 573). 86 In a debate in the National Assembly on citizenship in 1961, Nyerere defended a notion of “political citizenship” regardless of racial origin and called his racialist opponents, “little Hitlers” (cited in Iliffe 1979, 572). A law passed in 1965 allowed non-citizens to vote if have they have 5 years residence in the country (Mukandala interview). 87 McGowan & Wacirah 1974, 192 ff. The findings are based on sample of 705 elites covering the period from 1963 to 1968. 88 Mukandala, interview; Kaya 2004, 165. 89 Chris Peter, interview. 90 Bienen 1967, 268f. Some 40% of money income was derived from exports and 80% of these represented agricultural crops and livestock, with only about 13% earned from minerals. Three cash crops – coffee, cotton

Significantly, the means of production remained mostly in foreign hands since the process of transferring political power had not been accompanied by a corresponding process of transferring economic power. Also, and in sharp contrast to later periods, the public sector constituted only a limited reservoir of patronage. Even though TANU did distribute well-paid positions in government or party organs to maintain its inclusive coalition, Britain “set clear limitations on the speed of Africanisation of national institutions” and continued to dominate the major share of senior positions in the administration.91 As a consequence, the postcolonial government felt the urgent need to devise ways to create an economic base for the state, not least since this would also provide the material underpinnings for alimenting the “inclusive elite bargain”.
The imperative of economic reform was most obvious in the agricultural sector – the mainstay of the economy that was based on plantations, settlers and expanding (albeit still small) class of African capitalist farmers.92 Nyerere strongly disliked this agricultural structure as settler farms were mostly owned by foreigners93 and the trend towards African capitalist farming increased inequalities – a situation that clashed with his Socialist ideals and constituted a potential threat to TANU’s power base among rural peasants. There was thus urgent need for a new agricultural policy, not least since many settlers virtually abandoned large parts of the country upon independence. TANU opted for a policy designed to minimise socio-economic differentiation and inequalities94 – or in the words of Goran Hyden “small was strengthened at the expense of the large”.95
In 1962, the government decided “to nationalise all land in order to make the state the ultimate trustee of all land and rule out individual freeholding”.96 Even though the overall redistribution of land remained limited, small peasant producers were to be strengthened at the expense of large landowners, namely capitalist farmers (kulaks) and chiefs.97 This basic orientation was implemented through an agricultural policy based on two (somewhat contradictory) pillars. First, the government followed a “transformation approach” (based on recommendations in a World Bank report) that sought to modify peasant agriculture through the creation of settlement schemes.98 Theses schemes, which provided farmers (mainly TANU Youth) with land on condition that the followed pre-established rules were relatively few in number and mostly failed due to the use of inadequate technology. Second, and more importantly, the government adopted an “improvement approach” that was meant to enhance
                                                                                                                                                         and especially sisal – made up more than half of all export revenues. Subsistence activities were said to represent between one-third and two-fifths of GDP. 91 Harnevik 1993, 38. 92 Coulson 1982, 145ff. 93 “At independence, African ownership of the sisal plantations, a crop which constituted 50 per cent of exports, was 0.6 per cent, whereas Europeans held 75.3 per cent and Asians 24.1 per cent” (Klugman et al. 1999, 98). 94 This policy followed from TANU’s stance during the independence struggle. While the nationalist movement in Kenya fought for freehold rights, TANU – with its power base among rural peasants – resisted such policy as it “would have hastened rural differentiation and led to the emergence of a stratum of richer peasantry that the nationalist movement saw as unacceptable. Within the structure of the economy at that time, the number of richer African farmers who could have benefited from this law was extremely limited and it was perceived that the only group who would benefit from a capitalist land market were the European and Asian farmers at the expense of the non-capitalist peasants” (Khan & Gray 2006, 57 f.). 95 Hyden 1980, 82ff. 96 Hyden 1980, 70 97 In the Bukoba District, for instance, chiefs enjoyed traditional property rights over a portion of the land in their chiefdoms (through a semi-feudal land tenure system called “nyarubanja”). With the land reform of 1962, the “nyarubanja” was abolished and landlords lost their land to independent peasant producers (Hyden 1980, 82ff.). 98 Hyden 1980, 71ff.; Coulson 1982, 145ff.

the existing mode of peasant production.99 In order to increase the control of the economy by indigenous people, the cooperative movement was expanded in all parts of the country, including those previously untouched (central and coastal parts, Mtwara and Ruvuma in the south, and the western areas). Accordingly, the number of registered marketing cooperatives (for both export and food crops) increased from 857 in 1961 to 1533 in 1966. The expansion was based on “compulsory marketing order” whereby cooperatives became the only legal purchasers of crops. The agricultural extensive service was improved and channelled a large array of resources (inputs, credit, technology, training, education) to the cooperatives.
This agricultural policy favoured political stability in that the expansion of commercial agriculture in all parts of the country helped to reduce inequalities between ethnic groups and/ or regions.100 As the control over cooperatives provided local elites across the country with access to state resources, the expansion of the cooperative movement can be interpreted as a rent-sharing policy. At the same time, however, the government policy increased socioeconomic differentiation in the countryside. Even though it explicitly sought to reduce inequalities between commercial farmers and peasant producers, there is strong evidence that the expansion of the cooperative movement and the improvement of the agricultural extension service benefited mainly the large farmers – a situation that “allowed the strong to get stronger”.101 Altogether, agricultural policy during the early post-independence years mainly contributed to minimising “horizontal inequalities”, that is, inter-group inequalities that are arguably more conducive to large-scale violent conflict than interpersonal (or “vertical) inequalities.102 This “horizontal” inclusiveness was achieved “at a relatively low cost in terms of rent distribution” because “factional competition within the party structure was relatively limited compared to other developing countries”.103
In the industrial sector, the Tanganyikan government built on the late colonial policy of import-substitution.104 While there had been only very limited industrialisation under colonial role (cf. above), the country witnessed a relatively rapid process of import-substitution in light manufacturing between 1955 and 1965 (e.g. textile, sisal-spinning, chemical, plastic and rubber products, pre-cast concrete articles, cement, oil). This sudden “boom” became possible because Asians – whose traditional trading activities had been taken over by the cooperative movement – started investing in manufacturing, and because a number of companies established in Kenya (both Asian and multinational) decided to set up branches in Tanganyika (not least because Kenya became unstable after the Mau-Mau rebellion). In sharp contrast to the agricultural sector, however, the nascent industries remained overwhelmingly foreignowned – a situation would change only in the wake of the Arusha Declaration (see below).
SECURITY
Maybe more than any other single event, the military mutiny that occurred in 1964 and the response to it by the young TANU government had far reaching effects on long-term stability. The related mutiny in Uganda’s African Rifles, the mutiny by the Force Publique in Congo
                                                 99 Coulson 1982, 145ff. This “improvement approach” was not least motivated by the fact that TANU was anxious to take unpopular measures that could endanger its power base among rural producers. Rural tax collection, for instance, remained highly incomplete throughout the 1960s (Hyden 1980, 76ff.). 100 Harnevik 1993, 38. 101 Coulson 1982, 145ff. 102 For the distinction between „horizontal“ and „vertical“ inequalities and their varying impact on large-scale violent conflict cf. Stewart (2000). 103 Khan & Gray 2006, 40 104 Coulson 1982, 168ff.

and the different response to them may go a long way in demonstrating the distinct trajectories of state building in these countries.
At independence the units of the King’s Armed Rifles formed the core of the new country’s armed forces and they were under the command of British officers. The army was composed of 2,000 soldiers with only three African officers.105 Africanisation of the officer corps was proceeding slowly and soldiers faced persistent discrimination at the hands of British officers and developed grievances over low pay. Training was British oriented and did little to improve the skills of recruits. No universal conscription had yet been introduced. There was almost no contact between the political authorities of Tanganyika and the armed forces.106 As a consequence, the military “had no machinery of making its grievances known to the Government”.107 While TANU took care to accommodate other sectorial interests (e.g. the civil service or the trade unions), the army was ignored in the struggle for sharing the “fruits of independence”. This neglect of the armed forces can arguably explained by TANU’s confidence that it could keep “the army apolitical” and make it serve the civilian government.108
The outbreak of the military mutiny in 1964 shoed that TANU’s confidence was unfounded and shook the ruling party to its foundations. In a sense, since the mutiny lasted longer in Tanzania than in Uganda and involved almost the entire army, TANU felt more threatened by it than did Ugandan political authorities. As a result they decided to abolish the old army and construct an entirely new force. The new army, established in September 1964 and named the Tanzanian People’s Defence Force (TPDF) was subordinated to the ruling party and initially drew on the base of the TANU youth wing.109 While the number of recruited TANU youths remained too small and it soon became necessary to recall former members of the Tanganyika Rifles, TANU membership was required as a prerequisite to join the armed forces. Soldiers were recruited from all tribal origins110 and different walks of life with promotions based on performance rather than affinity. The President retained the power to make appointments from the chief of the TPDF down to battalion commanders. Political commissars were introduced into the army in order to ensure “a new, correctly politicised military establishment”, and senior officers were awarded prominent party positions.111 Some 25% of training time was reserved for political study, which stressed the history of the national struggle and the ideology of the party. TANU and Nyerere consciously worked to create an idea of the TPDF as a “people’s army”. In 1965 national service was introduced, eventually becoming voluntary for primary school leavers, but obligatory for secondary leavers for two years. In order to ensure that the politicisation and popularisation of the army would not undermine its professionalism, a national military academy was set up and sources of training diversified.112
                                                 105 In fact, the British had introduced the first African political officer in 1955 (Iliffe 1979, 573). 106 Lupogo 2001.  107 Swai 1991, 94f. 108 Omari 2002, 93. When asked whether the military mutiny in Congo could be repeated in Tanganyika after independence, Nyerere even overconfidently replied that “[t]hese things cannot happen here. We have a strong organisation, TANU. The Congo did not have that kind of organisation. And further there is not the slightest chance that forces of law and order in Tanganyika will mutiny” (cited after: ibid.). 109 Omari 2002, 94. 110 Zirker (1992, 112) reports that a “strict formula” of balanced ethnic recruitment and promotion was introduced after 1964. See also Lupogo 2001. 111 Zirker 1992, 112. 112 Omari 2002, 94. Israel, Nigeria, Canada, China, India and North Korea. Israel, however, concentrated on the National Service.

In 1968 Nyerere understood just how important the mutiny had been to state-making in the country:
“The mutiny was a strike of the army people, and it went out of control. It shocked the country. But every cloud has a silver lining, as the British say. It enabled us to build an army almost from scratch. Many institutions we have inherited, but the army is something we built ourselves."113
The complete reconstitution of the armed forces in Tanzania and the fact that they were integrated into TANU and made subservient to political authority was crucial to the state’s stability in years of crisis ahead. This move probably consolidated the resolve to maintain TANU as a single party, since such political control could not be achieved over the armed forces in a multiparty arrangement.114

IV Ujamaa and nationalisations (1967-1981)  

By the mid-1960s, the Tanzanian state started to face a credibility crisis. Annual GDP growth rate of 6% during the early post-independence years notwithstanding, Nyerere himself evaluated the country’s economic performance as “growth without development”, largely insufficient to bring about the impatiently expected improvements in material welfare.115 Control over the national economy still remained limited thereby constraining the capacity to implement redistributive policies. Moreover, “there were manifestations of a growing divide between the TANU leaders and the bureaucracy on the one hand and the social base of the nationalist movement, poor peasants and workers on the other”.116 While the social profile of post-colonial policies had been neglected, socio-economic differentiation was on the rise with elites struggling for “status, income, and personal power”. The overall situation was further complicated by the inability to realise sufficient foreign aid, which was mainly due to a diplomatic fallout with Britain, the United States and West Germany.
The Arusha Declaration (1967) has to be understood as a response to these mounting contradictions within the post-colonial state. The declaration spelled out Nyerere’s visions of a socialist society, emphasising public ownership, self-reliance as well as the importance of agriculture and rural development. It was followed by radical transformative policies – most notably large-scale nationalisation and villagisation – that were (unsurprisingly) accompanied by social tensions and incidences of minor-scale violence. In 1969, for instance, a group of army officers and disgruntled politicians, whose property had been nationalised, launched an unsuccessful coup attempt.117 Also, the compulsory villagisation scheme during the 1970s often involved a significant use of state force, most notably against the owners of cash-crop producing land.118 Altogether, however, the Tanzanian state showed remarkable resilience and again remained free of large-scale violent conflict. In what follows, we argue that this enduring political stability can be explained by the ruling party’s ability to maintain and even extend its centralised and inclusive authority in both the civilian and military sphere. Even more importantly, the Arusha policies provided the post-colonial government with tight control over the economy and thereby established the material basis of the “elite bargain”.
                                                 113 Lupogo 2001. 114 Omari 2002, 94. 115 Nyerere 1968, cited after Harnevik et al. 1988, 95 116 Harnevik 1993, 40f. 117 Lupogo 2001. 118 A few examples for this state-sponsored violence are documented in Coulson 1982, 250ff.

POLITICS
Between 1967 and 1981, TANU undertook further measures to consolidate its centralised authority whereby it came to enjoy an even tighter grip over both the Tanzanian elite and the broader population.
The TANU leadership code (1967) was a key measure to gain control over elites and contain private enrichment and corruption. It required all state elites to undergo political training by the ruling party and banned them from owning more than one real property, investing in business and receiving a salary outside their state employment. According to Costello, the code was not least an attempt to better incorporate civil servants into the party-state.119 As elite cohesion had suffered from persistent competition and disagreements between party and administration throughout the 1960s, TANU sought to make bureaucrats more dependent on the state and ensure “compliance with party politics”. In a similar vein, the TANU Guidelines (Mwongozo) in 1971 were supposed to impose party control by creating party branches in all national institutions, including the proliferating parastatals.120 Hyden claims that this was mainly “to ensure greater control of a faction of the petty-bourgeoisie” that benefited from their position in the parastatals and had proved disloyal to political leaders in other African countries. Altogether, no leader “could any longer claim autonomy from the party control” whereby the cohesion of the “elite bargain” had been improved.
Party control was also imposed at the regional and local level where elites had been able to preserve a considerable degree of autonomy throughout the 1960s (see above). 121 In a first step, TANU initiated a decentralisation programme (1972) that was rather a “centralisation” in that it replaced the elected and self-reliant District Councils with centrally appointed Area und Regional Commissioners.122 As the District Councils had long been perceived as instruments of local notables, their suppression allowed TANU to restrict the potential for independent political mobilisation at the local level.123 Furthermore, the decentralisation of the government administration anchored planning and implementation at the regional and district levels and increased the state’s control over rural areas. At the same time, TANU underwent a similar reorganisation. As salaried officials were strengthened at the expense of elected ones, there was for the first time a “strong, well-paid party bureaucracy at regional, district, divisional, and ward level”. 124 The ruling party thereby became the de facto instrument of local government. In a second step, TANU decided to abolish the cooperative unions in 1976. The cooperatives were typically controlled by larger farmers (kulaks) who had a local power base, controlled a large share of the exchanges but were not fully integrated into the ruling party.125 Similar to the elected District Councils, they therefore constituted fairly autonomous structures that TANU found recurrently difficult to control. As a consequence, the cooperatives were replaced by parastatal companies (“Corp Authorities”) that were now dealing directly with the sale and purchase of the crops. Once local elites had been severely weakened, bureaucrats and party cadres were able to develop centralised “patronage relations” with villages. A key instrument in this regard was the regional development funds that were introduced in 1976 and exclusively devoted to the villages. These funds enabled bureaucrats or party cadres to better compete with local elites who had
                                                 119 Costello 1996, 124ff. 120 Hyden 1980, 159ff. 121 Hyden 1980, 107 122 Coulson 1982, 254 123 Hyden 1980, 134ff. 124 Coulson 1982, 254 125 Hyden 1980, 132ff.

previously controlled patronage – the party patron was strengthened at the expense of the local patron. Altogether, the state became more directly anchored at the grassroots and thereby extended its control over rural areas.
Having established control over virtually all parts of the polity, TANU formally entrenched its supremacy in the Interim Constitution of 1975 that declared it the supreme organ of the state.126 The 1977 Constitution made the Party, not the Cabinet, the main organ for advising the President, even though the Cabinet continued to exercise a key advisory function because of its superior technical know-how. Also, the ruling party’s machinery was further strengthened when TANU decided to merge with the Afro-Shirazi Party (ASP) in Zanzibar in 1977. This saw the birth of the Chama Cha Mapinduzi (CCM) – or “Revolutionary Party” – which has remained the ruling party ever since.
Beyond the consolidation of centralised authority, TANU also took great care to maintain an inclusive elite coalition. With respect to the staffing of his Cabinet, Nyerere repeatedly made gestures in “the direction of regional and religious balancing”.127 While the distribution of public employment was overall fairly balanced, there was – as in the early post-independence period – evidence for regional and religious imbalances. As public employment required educational qualification, the North benefited more than the less educated South and the better educated Christians outnumbered Muslims.
ECONOMY
The Arusha declaration involved the nationalisation of large parts of the Tanzanian economy and the proliferation of parastatals in almost every sector. This increased share of the state in the economy marked “an important turning point in the nature of rent creation by the state and the associated rent seeking”.128 According to Coulson, the parastatal was to perform three fundamental roles, namely (1) reducing the transfer of profits out of the country, (2) expanding investment in productive sectors, and (3) strengthening infrastructure. 129 Altogether, the number of parastatals rose from 64 in 1967 to 380 in 1981.130 In the manufacturing sector, the state’s share rose from 14% in 1967 to 57% in 1982.131 Furthermore, “extensive administrative resource allocation and price controls, import quotas and foreign exchange rationing system were introduced”.132
The motivations underlying the Arusha policies were at least twofold. To begin with, its socialist rhetoric of minimising inequalities and re-establishing unity in society should not simply be discarded as done by some leftist critiques.133 In this sense, Harnevik is maybe right                                                  126 Harnevik et al. 1988, 101. 127 Kelsall 2002, 608f. 128 Khan & Gray 2006, 41. 129 Coulson 1982, 274ff. Key financial parastatals included the Bank of Tanzania, the National Bank of Commerce, the National Insurance Corporation and the State Trading Corporation, whereas manufacturing parastatals comprised investment banks (e.g. the Tanzania Rural Development Bank), holding companies (mainly the National Development Corporation) as well as parastatals directly involved in production. Key infrastructure parastatals, finally, involved railways, harbours, posts and telecommunications. 130 Harnevik 1993, 50. When compared with other African countries, this number highlights the exceptional degree of state involvement achieved in Tanzania. By the 1980s, only Mozambique had a larger number of parastatals with over 1000. Senegal had 188, Ethiopia 180, Ghana 130, Guinea 181, and the majority including Kenya, Uganda and Zambia had fewer than 100 (Temu & Due 2000, 693). 131 Klugman et al. 1999, 84. 132 Khan & Gray 2006, 41. 133 E.g. Shivji 1976 who tends to reduce the goal of the declaration to providing the “bureaucratic bourgeoisie” with an economic basis.

to argue that the “primary aim was to change the direction of societal development so that the TANU leadership, the bureaucrats and the social base of the nationalist movement would grow closer together”.134 From this perspective, the goal was to achieve rapid economic growth “without the unequalising pressures of capitalist development” that would have mainly benefited Asian and European business interests as well as the small class of African capitalist farmers.135 Nonetheless, and this is equally important, the extremely high degree of state intervention in the economy was also motivated by the perceived need to create a material base for Tanzania’s “inclusive elite bargain”. The expanded parastatal sector brought an end to the predominance of European and Asian interests and created – together with the steadily increasing employment in government and party organs – a huge reservoir of wellpaid employment that made it relatively easy to satisfy the country’s educated elite and accommodate ethnic, regional and religious cleavages. These far-reaching rent-sharing policies created a high degree of elite unity and thereby contributed to political stability.
The agricultural sector, the backbone of Nyerere’s Tanzanian socialism, was subject to particularly transformative forms of state intervention. The government launched the concept of ujamma (literally: familyhood), aiming at the “creation of communal village production units”.136 This policy was explicitly directed against the large farmers and sought so achieve a more equitable distribution of wealth – a goal that had not been achieved in the early postindependence period (see above). From 1970/71, TANU started to assume an active role in village creation, which involved explicit campaigns against capitalist farming. As the number of villages grew from 1956 in 1970 to 5631 in 1973 (sheltering around 2 million people), European capitalist farmers (notably in Arusha and Kilimanjaro Regions) began to leave the country and most of capitalist farming came to an end. While half of the sisal plantations were nationalised, the coffee estates in Kilimanjaro as the last capitalist enterprises were handed over to local cooperatives in 1974. Between 1973 and 1976, the government intensified its efforts to contain the emerging African capitalism. The compulsory villagisation policy was the “largest resettlement effort in the history of Africa” with about five millions rural Tanzanians being resettled.137 By 1977, there were 7684 villages involving a total of 13 million people.138 This radical resettlement effort, which involved an element of land reform, combined with the subsequent dissolution of the cooperative unions in 1976 (see above) were decisive measures to promote a more equitable distribution of public goods while increasing state control over economic activity. Their combined effect further undermined the power of the richer peasants.
While the political impact of these policies remains difficult to evaluate, there is reason to argue that the established agricultural order was rather conducive to political stability. First, large-scale villagisation combined with the introduction of state marketing structures provided the ruling coalition with tight control over the agricultural surplus – the main source of wealth in the country – and thereby represented a cornerstone of its hegemony. The transfer of financial surplus from the peasantry to the state was considerable throughout the 1970s amounting to an estimated average tax of 26.6% on peasant crop income.139 Second, the villages were created in all parts of the country whereby all regions and ethnic groups benefited from access to state resources. Also, and in contradiction with the anti-elite bias of agricultural policies, there is evidence that local elites typically benefited disproportionately
                                                 134 Harnevik 1993, 43. 135 Khan & Gray 2006, 41 136 Hyden 1980, 96ff. 137 Hyden 1980, 129ff. 138 Coulson 1982, 242. 139 Ellis 1983, cited after Harnevik 1993, 48.

from villagisation policies140 whereby they had little incentive to mobilise opposition. Third, although regional inequities continued to be obvious or maybe even increased141, they were “compensated” as a result of a deliberate policy to extend the social infrastructure primarily to the more backward parts of the country”.142 These service delivery programmes were an integral part of ujamma policies, the argument being that people had to live in villages in order for the government to reach them. Progress in providing basic education and health services was immense and improved the social status of the broader population – a situation that secured the government “wide-based political support”.143 TANU “established very successful universal primary education and adult education programmes and opened health centres throughout the country rather than large hospitals accessible only by a few. During 1971-74, education was the single largest item (16 per cent) in total government spending. As a result, literacy increased from 10 per cent in 1960 to 30 per cent in 1970 and 70 per cent by the end of the 1970s. Life expectancy rose and child mortality fell substantially.”144
Some people have even argued that relocations through villagisation “may have weakened tribal connections“thereby exercising a positive impact on national integration.145 While this seems plausible at first sight, empirical evidence does not support such argument. While millions of Tanzanians were relocated (see above), the newly created villages were on average no more than 5km away from the original dwellings thereby minimising the potential for ethnic blending.146 Also, Hyden reports that in some instances even “fighting broke out between members of the Nyaturu and the Barabaig tribes over land rights that were upset in conjunction with the enforced movement of people into villages”.147
SECURITY
A series of events in the late 1960s and early 1970s strengthened TANU’s resolve to further incorporate the army into the ruling party. First, the failed coup plot of 1969 (see above) illustrated that the military still represented a reasonable threat to civilian rule. Also, the Portuguese invasion of Guinea in 1971 showed just how vulnerable African states could be to outside interference.148 Even more importantly, the fall of Milton Obote’s government to Idi Amin’s coup d’etat in Uganda, sent warning signals throughout the TANU leadership. Finally, on Zanzibar, Sheikh Abeid Amani Karume was assassinated in 1972 and some thought it was at the hands of military officers.
As a consequence, the 1971 Party Guidelines also redefined the relationship between the ruling party and the armed forces. While all soldiers had been required to join TANU from 1964 (see above), it was not until after 1971 that a party structure was introduced into the army.149 This party structure in the military – operated from 1971 and approved by the TANU National Executive Committee (NEC) in 1978 – was modelled after the overall structure of
                                                 140 Coulson 1982, 246. 141 There is some evidence for widening differences in per-capita income between regions. The cash income of smallholders in the rich regions (Tanga, Kilimanjaro, Arusha) hardly fell between 1969 and 1975, but middleincome regions (Kigoma, Mtwara, Lindi, Mwanza, Ruvuma, Coast, and Dodoma) lost nearly 20 per cent of their cash income, while the poor regions lost nearly half (Coulson 1982, 197). 142 Hyden 1980, 122. 143 Klugman et al. 1999, 85. 144 Klugman et al. 1999, 85. 145 Klugman et al. 1999, 80. 146 Martin 1988, 111. 147 Hyden 1980, 151. 148 Omari 2002, 96ff. 149 Swai 1991, 100f.; Omari 2002, 100ff.

the ruling party and introduced a “kind of democratic centralism” in the army”. While the army party hierarchy remained subordinate to the national party organ (that is, the NEC), it nonetheless gave the military an opportunity to participate in the formulation of Party Policy. This had considerable politicisation and identification effects and ultimately eased civilmilitary relations.
Beyond this “bureaucratisation of the military”, measures were taken to further the integration of the army into the civil service.150 From 1971, military personnel were increasingly co-opted into civilian sphere (as managers, directors and to party and government posts). Moreover, civil servants and military officers were allowed to run for political office in 1972 without having to give up their jobs. In case of election, they were granted an unpaid leave and could return to their original position upon completion of their service. This progressive “militarization of the bureaucracy” gained pace throughout the 1970s and especially the 1980s (see below).
The increasing integration between the ruling party and the army was from the early 1970s complemented by the creation and training of a militia, drawing on traditions of village selfdefence. The recruitment of the militia was done by the Party (but recruits were not limited to TANU members only), while the army carried out the training.151 Upon complete of the training recruits became part of a “reserve army” whose command was directly under the Party rather than the military. This “militarization of society” did not only build confidence between the broader population and the army (that had been perceived as an alien and oppressive force since colonial rule) but also provided TANU with a counterbalance to the TPDF, which no longer monopolised the use of force.
The war with Uganda at the end of the 1970s had an ambiguous impact on civil-military relations. On the one hand, it demonstrated Tanzania’s achievements in building state resilience over the past twenty years. As Uganda’s armed forces (under Amin) invaded Tanzania’s Kagera Region in northwest in November 1978, the Nyerere government immediately mobilised for a counter-attack calling on and apparently receiving popular support to fight the Amin dictatorship.152 According to Lupogo, in just a few weeks, the armed forces were expanded from 40,000 to 100,000 fighters, by drawing on the police, prison services and militias. The speed and potency of the response was unprecedented and led to a sweeping victory by April 1979. On the other hand, however, the size of the campaign and the impact it had on the armed forces generated new problems. As Tanzanian forces returned victorious there were no immediate efforts to downsize the army, which had grown enormously due to the war effort. The military was virtually given a free reign to establish these units at home with little regard to the costs this would incur on the state budget or the economy, particularly given the context of economic crisis in the country. Militias, expanded for the war effort, got involved with robbery and other criminal activities. 153 As a consequence, TANU had to devise further measures to ensure the subordination of the military throughout the 1980s (see below).

V Economic crisis and liberalisation (1981-1991)  

                                                 150 Omari 2002, 101ff. 151 Swai 1991, 96 152 Lupogo 2001. 153 Shivji 1990, p.144. To be added to the bibliography.

In the early 1980s, Tanzania was hit by economic crisis that was dramatic even by African standards.154 It is now widely agreed that the downturn was the result of both internal and external factors, including currency overvaluation, low producer prices, parastatal mismanagement, declining terms of trade, the oil crises of 1974 and 1978, the droughts of the mid-1970s, the break-up of the East African Community (EAC) in 1977 and the war with Uganda between 1977 and 1979. As a result of these factors, GDP per capita declined by an average of 1.67% between 1980 and 1986 (thereby falling to a level even lower than in 1966),155 while inflation reached around 30% throughout the 1980s. The budget deficit deteriorated considerably, averaging 10.09% between 1980 and 1986. All critical export crops, except coffee and tea, suffered from disastrous decreases.156 The ensuing acute balance-ofpayments crisis resulted in a growing debt crisis (the estimated debt service ration climbing to staggering 66% by the mid 1980s) and extreme scarcities of foreign exchange. As funds for the funds for the importation of capital goods, spare parts, and raw materials became scarce, industrial production plummeted by 11.3% between 1980 and 1985 (industrial capacity use of about 20%). Also, the availability of consumer goods and the provision of social services and infrastructure were very negatively affected. Economic stagnation saw the expansion of an “informal economy”, which came to be of far greater dimensions than the formal economy.157
This “worst of all worlds”158 was of course not free of political and socio-economic tensions. From 1983, for instance, the country witnessed an increasing degree of state repression against alleged “economic saboteurs”. In 1986, thousands of workers demonstrated because they had not been paid. Instead of paying the workers, the management called in the Field Force unit – a paramilitary force that opened fire on the unarmed worked, killing four and injuring many others.159 Also, there was once more a failed coup d’Etat in 1982-83 that involved about 20 army officers and a few civilians, largely organised by Tanzanians living abroad.160 These incidences notwithstanding, the country once again avoided extremer forms of large-scale violent conflict. This political stability against the odds can be explained by CCM’s ability to reconfigure itself in times of crisis without endangering its integrity, while preserving both the material basis of the elite bargain and control over the armed forces.
POLITICS
Since independence, the ruling party had managed to achieve a fairly high degree of legitimacy through its nationalist rhetoric, the improvement of social services and the profile of Nyerere as a credible and committed leader. Against the backdrop of dramatic economic downturn in the early 1980s, however, CCM started to face a severe crisis of political legitimacy.161 Unable to provide rapid solutions to economic turmoil and decline, the ruling party’s ability to fulfil the promises of the Arusha Declaration was brought into question – a situation that weakened its mobilisation capacity. Mounting unemployment, especially among urban youth, and a lack of opportunities for income generation constituted an increasing threat
                                                 154 The following account of economic crisis is based on Lofchie 1993, 418ff. and Klugman et al. 1999, 87.  155 Despite a relative recovery from 1987, Tanzania’s GDP per capita in 1999 was “the second lowest in the world, exceeding only Mozambique, which had been devastated by more than two decades of civil war” (Lofchie 1993, 418). 156 During the period 1980-84, cotton production fell to 65% of its 1970-72 peak; sisal production – Tanzania’s key export in the 1960s – to 28% of its 1964-1966 peak, and cashew production – Tanzania’s key export in the 1970s to 30% of its 1972-74 peak (Lofchie 1993: 419). 157 On Tanzania’s informal economy cf. Maliyamkono & Bagachwa 1990. 158 Lofchie 1993, 421 159 Campbell 1992, 100 160 Swai 1991, 113f.  161 Baregu 1994, 165ff.

of political unrest. Economic crisis was therefore not merely an economic problem but also began to undermine CCM’s political achievements.
CCM’s response to waning political legitimacy was threefold. First, the ruling party adopted a repressive stance that continued the increasingly authoritarian tendencies of the late 1970s.162 While the Economic Sabotage Act (1983) was officially presented as a mere economic action to combat hoarding, overpricing and others kinds of economic sabotage, Maliyamkono & Bagacliwa (1990) argue that it was not least “prompted by the perceived need to counter a threat to state legitimacy”.163 Under this Act, special tribunals allowed neither for legal representation nor for bail, all rules of evidence were set aside and police was granted extensive powers of investigation, search and detention. Moreover, the concomitant Human Resources Deployment Act (1983) gave the government considerable leeway in addressing the problem of urban unemployment, e.g. by means of forced relocations. On the one hand, these acts created the popular impression that not only government but also private actors were responsible for economic havoc, which may have helped to contain unrest. On the other hand, they also further undermined the government’s political legitimacy, especially when it became obvious that the state would not use the acts against higher state elites (see below).
Second, CCM soon engaged in a (albeit limited and controlled) reduction of centralised authority that it had so carefully built throughout the 1960s and 1970s.164 In 1982, party and government were separated at the regional and district level, which gave rise to greater autonomy and influence for local commercial elites. The latter also benefited from the reintroduction of the cooperative unions that replaced state crop authorities during the same year.165 However, the cooperative movement was as autonomous as in the 1960s since CCM took great care to control its organisation at the top. As primary societies were re-registered and started to operate from 1984, it soon became clear that their organisational structure extending over several villages – which was economically desirable – constituted a threat to the Party-dominated village political structure. This clash between economic and political objectives was resolved by a Party directive in 1987 that limited primary societies to one village only. This move underlines CCM’s continuing preoccupation with political control at the expense of economic efficiency.
Third, and clearly most importantly, the ruling coalition proved mature enough to reconfigure itself in times of crisis without endangering its integrity. Significantly, economic crisis gave rise to a power shift both within CCM and among the central organs of the state. Throughout the 1980s, a sharp division developed within Tanzania’s political leadership between those in favour of economic reform (the “pragmatists”) and those who remained committed to the old statist system (the “socialists”).166 This division conformed roughly to the division between the state administration on the one hand, and CCM on the other (even though pragmatists and socialists were found in both the administration and CCM). As CCM remained the supreme organ of government, the pragmatists faced the challenge of dislodging the party’s most prominent socialists from important ministerial positions, while avoiding a destabilising rupture between the party and the state. According to Costello, when economic crisis undermined the capacity and legitimacy of the party, bureaucrats – who had never fully been incorporated into the party (see above) and were now bolstered by their privileged ties to foreign actors (especially the IMF) – “became an increasingly coherent and dominant force                                                  162 Harnevik 1993, 58 163 Maliyamkono & Bagacliwa 1990, cited after Costello 1996, 138 164 Harnevik 1993, 61 165 Harnevik et al. 1988, 61ff. 166 Lofchie 1993, 452f.

within the state”.167 The rising influence of the pragmatists was behind Nyerere’s resignation as President in 1985 who now dedicated his attention – as chair of the CCM – to reconstructing the link between the party and the people.168 Ali Hassan Mwinyi's ascension as the new President brought to power a “career bureaucrat, whose power was based less in the party than in the government”.169 Mwinyi consistently sought to strengthen the government at the expense of the ruling party, particularly with respect to policy making.170 While the socialist faction no longer controlled the key organs of government within which economic policies were formulated, CCM nonetheless continued to maintain a “high political profile” (e.g. in Parliament), which caused uneasy coexistence of a reformist government and a socialist party.171 This situation was resolved in August 1990 when Mwinyi became chairman of the CCM as well as President. Leading government and party positions were again reconciled, but now under the control of the reform-oriented pragmatic faction.172
Altogether, CCM managed to maintain a high degree of centralised authority during economic crisis, albeit under new leadership and with increased leeway for the state bureaucracy. Moreover, there is no evidence that economic crisis put an end to the inclusive character of ruling party. While ethnic and regional favouritism continued to be avoided, President Mwinyi was “widely perceived to have gone some way to redressing religious imbalance by promoting Muslims in his government”.173
ECONOMY
The dramatic economic crisis of the 1980s represented an imminent threat to the material base of Tanzania’s elite bargain. As the crisis-ridden economy generated a sharply declining supply of rents to bind together and reproduce the ruling political coalition, there was a clearly perceived need for economic adjustment and reform – not least since even middle and upper class civil servants were seriously affected by dwindling purchasing power and the extreme unavailability of even the most basic consumer goods.174 However, as Stein as convincingly argued, the debate during ensuing process of economic adjustment was about changes in policy rather than in elite power and did therefore “not challenge the existing hegemony of the state”.175 Instead, Tanzania’s “inclusive elite bargain” remained largely intact, not least at the expenses of the living standards of the bulk of the population.176
                                                 167 Costello 1996, 123 168 By resigning from the presidency but remaining at the helm of the CCM, Nyerere was able to keep up a critical stance towards the IMF, which helped to preserve the legitimacy of the CCM and ultimately the political alliance in control of the state (Stein 1990, 20). To be added to the bibliography. 169 Costello 1996, 138ff. 170 In 1988, for instance, “Mwinyi issued a circular stating that bureaucrats who wished to earn a second income during their own time could do so, as long as they gave the state the forty hours a week for which they were paid. In one step, this abrogated the Leadership Code; only in 1990 did the party open this issue for discussion. The president had presented the CCM with a fait accompli, the Leadership Code was passé, no matter what the party said. The government's initiation of this change demonstrates the power to define policy was shifting from party to government” (Costello 1996, 140, original emphasis). 171 Lofchie 1993, 454 172 Costello 1996, 145 173 Kelsall 2002, 610 174 Lofchie 1993, 447 175 Stein 1992, 59f.  176 This closely mirrors patterns of structural adjustment across the whole of Sub-Saharan Africa as described by Nicholas van de Walle (2001).

Economic reform in Tanzania was characterised by difficult negotiations with the IMF and the World Bank.177 While the initial IMF adjustment package (1981) was – under the vehement leadership of Nyerere – rejected as incompatible with the Tanzanian development strategy, the period between 1981 and 1986 was nonetheless used to prepare for and “own” an accommodation with the international financial institutions, both in political and economic terms. In late 1981, the government first launched the National Economic Survival Program (NESP) – a soon abandoned “home grown” reform effort that suffered from unrealistic (export) targets and failed to mobilise internal resources. The 1982 Structural Adjustment Programme (SAP) combined attempts to preserve the parastatal sector and the social profile with the cautious introduction of liberal reform elements (increased producer prices, reductions in consumer subsidies and a limited devaluation) – an approach that did not find the approval of international donors. In the same year, the National Agriculture Policy (NAP) opened the way for elements of private ownership of land and advocated the liberalisation of various aspects of agricultural marketing, while still supporting producer subsidies and the use of existing state-controlled marketing boards. The real liberal shift came however with the 1984/85 budget that foresaw, among others, a significant raise in producer prices, a large currency devaluation, the removal of subsidies on maize meal and fertilisers and the introduction of a “own funded import scheme” where no questions were asked about the source of foreign exchange. In the light of these previous policies, Tanzania finally reached an agreement with the IMF in August 1986 and was able to present the accord as a logical extension of its own efforts. Major policies announced in the Economic Recovery Programme (ERP) included a series of further devaluations; an increase of producer prices by 5% per year, the further liberalisation of external and internal trade by tightening budget expenditure and money supply; and improvements in parastatal efficiency and foreign exchange allocation. Upon expiry of the ERP, a second phase of structural adjustment was initiated covering the period from 1989-1989. The “Economic and Social Action Programme” (ESAP) mainly sought to build on previous trade liberalisation targeting the liberalisation of foreign investment, the financial sector, and agricultural marketing.
Economic adjustment was accompanied by deteriorating living standards for the bulk of the population. Due to persistently high levels of inflation during the 1980s (see above), real urban wages declined by 63% between 1981 and 1989, while the removal of maize subsidies (the urban staple food) caused further hardship.178 The rural population seems to have fared slightly better at first sight since real producer prices increased by an average of 24% between 1980/81 and 1987/88. If one excludes cashew nuts and robusta coffee (which have outlier status), however, price increases are reduced to a negligible 3,9%. In some cases, devaluationinduced-inflation and rising input costs even by far outweighed the nominal increases in producer prices.179 Furthermore, the state clearly used the devaluations to extract a larger percentage of the values of the crops. Between, 1986-1983, for example, the peasants’ share of the value of major export crops dropped from around 80% to 50%.180 Finally, and perhaps most importantly, the adjustment process involved grossly deteriorating social services, especially for the poor. While the share of education in total government spending declined from 12.55% to 6.45% between 1980 and 1987, the share of health fell from 5.61% to 3.66% over the same period (see table XXX). As school fees were introduced in 1984, the literacy rate dropped by 7% among the poor but increased among the better-off between 1983 and
                                                 177 The following account of the adjustment process is based on Stein 1992, 65ff. and Harnevik 1993, 287ff. 178 Stein 1992, 65ff. 179 While the average producer price for cotton increased by 70% between 1985 and 1989, the average prices of agrochemicals for cotton went up by 475% during the same period. This amounted to real price decrease of 14.6% (Harnevik 1993, 298). 180 Harnevik 1993, 298
 27
1991.181 Meanwhile, 6% of the better-off had access to secondary school, whereas only 1% of the poor had this privilege. By the early 1990s, only in two of the 20 regions (Arusha and Kilimanjaro), did more than 10% of the population above school age have secondary or higher education. Altogether, “the government’s objective of eliminating regional biases had become severely constrained by the 1980s”.182
While CCM leadership was asking ordinary people to tighten their belts (with slogans like hali halisi – life is hard), the higher levels of the ruling elite maintained its conspicuous consumption patterns.183 In the 1980s, budget allocations to the public service increased almost steadily– despite the dramatic economic crisis and a concomitant decrease in the expenditures for education and health (see table XXX).

Table XXX: Selected government expenditure by purpose (%)184 

Structural reform in public administration and the parastatal sector – the bastions of Tanzania’s elite bargain – remained minimal throughout the 1980s. After a commitment to reduce the size of the government, the number of cabinet ministers was reduced from 40 to 30 in 1985 but then crept back to 38 by 1990.185 Civil service reforms started as early as 1982 but did not lead to noticeable results until the mid-1990s. In fact, total government employment under Mwinyi continued to grow until 1992, when it reached a peak of 355,000 employees.186 Furthermore, the parastatal sector remained largely untouched. In 1988, Tanzania still had 413 public enterprises, of which 339 were commercial, and 56 non-commercial (e.g. higher learning institutions), plus several holding companies.187 Thus, even after committing to the liberalization in 1985, the parastatal sector continued to grow and direct subsidies to parastatals increased to 11% of total government expenditure during the late 1980s.188
This is not to deny that elites were also adversely affected by economic crisis. The purchasing power of the civil service, for instance, declined by 94% in real terms between 1969 and 1985.189 Yet, various forms of corruption that became rampant from the early 1980s opened new ways for elite accumulation. As the formal economy declined, civil servants started to make use of their control over a wide array of licences, taxes, subsidies, quotas, and rationing
                                                 181 Klugman et al. 1999, 91 182 Ibid. 183 Campbell 1992, 97 184 Campbell 1992, 92f. 185 Kiondo 1992, 32f. 186 Kjaer 2004: 399 ff. 187 Temu & Due 2000, 693. 188 Khan & Gray 2006: 64. 189 Stevens 1994, cited after van de Walle 2001, 155.

mechanisms to derive extra income. 190 Significantly, the highly partial economic reform adopted since 1985 (see above) has by no means contained these illicit economic activities, which appear to have continued to grow. It was especially after Tanzania embarked on the process of partial market liberalisation in the mid-1980s that “the one-party state began to resemble a racket for the protection of corrupt and acquisitive public officials”.191 Tellingly, these forms of economic sabotage by middle- and higher level elites were hardly targeted by the above-mentioned Anti-Economic Sabotage Act (1983).192 As a result of epidemic of corruption, land grabbing, and lawlessness, Mwinyi's rule is often “remembered as a period of ruksa, a Swahili word perhaps best translated as ‘do your own thing”.193
SECURITY
At the beginning of the 1980s, CCM felt an urgent need to further subordinate the military under civilian control. The war with Uganda left the country with the potentially dangerous legacy of an oversized and increasingly confident army that became difficult to control. The failed coup attempt of 1982/83 was a timely warning signal in the respect. Moreover, the serious economic downturn increased the overall vulnerability of the ruling coalition and made the military back-up all the more necessary.
In 1981, the Mwongozo wa CCM for the first time granted the military formal representation at the ruling party’s sittings.194 This meant that the armed forces were turned into a “Party Region” with “districts” and therefore got the right to elect their own representatives to the National Conference. The military representation was equivalent to three districts – a non- negligible representation if one considers that the entire Party had no more than 113 in the country. The “military region” is chaired by the chief of the defence forces who by virtue of that position becomes a member of the NEC and the National Conference.
Furthermore, military personnel were increasingly co-opted into political leadership positions. From 1982, five seats of the CCM NEC were set aside to be contested exclusively by the army.195 Cabinet posts were also open to army officers either through election to parliament or direct presidential appointment.196 In 1982, more that 30% of all regional party secretaries were military personnel, while 15% of the district party secretaries and commissioners were army officers.197 Five years later, 15% of the central committee were senior officers, whereas 24% of the regional party secretaries and 2% of the NEC were former soldiers. By 1992, “the Tanzanian administration looked like a civilian-military coalition”.198 “By 1992 the Tanzanian administration looked like a civilian-military coalition.” As a consequence, “the armed forces did not feel left out of the action (…). They could therefore not point accusing fingers at other political leaders because they were in it together. The military was part of the government and the party hierarchy. It could almost be maintained that they did not need a coup d’état.”199
                                                 190 Tripp 1997, cited after van de Walle 2001, 154. 191 Kelsall 2003, 56 192 Stein 1992, 65ff.; Harnevik 1993, 58. Instead, the Act targeted mainly „small fry“ as 92% of the arrested were private businessmen and the rest low-level officials. 193 Hyden 1999, 143. 194 Swai 1991, 100ff. 195 Swai 1991, 101 196 Swai (1991, 101) reports that Members of Parliament with military background increased from 4 between 1975-1980 to 22 between 1982-1987. 197 Omari 2002, 101 198 Omari 2002, 101 199 Lupogo 2001.

The critical importance accorded to the military was also reflected in the rising military budgets, which – according to Zirker – “appear to represent a systematic strategy of cooptation”.200 Between 1970 and 1990, the percentage of the national budget allocated to the military increased from 4% to 14%, while the budget share for education decreased from 14% to 4% during the same period.201
Beyond the co-optation of the armed forces, CCM also managed to contain the Sungu Sungu movement that developed in the most populous regions of the country (Mwanza, Shinyanga and Tabora) “as a popular peasant army reacting to the cattle rustling and illegal smuggling of gold and ivory from the rural areas”.202 As the state failed to deliver security to ordinary citizens, especially in the countryside, the movement started to dispense justice itself through popular democratic village assemblies – a form of self-organisation that bypassed the courts, the police and the party structures. Faced with this “institutional multiplicity”, the government initially “deployed the security organs to infiltrate the movement, but in the face of its mammoth growth to over 5 million citizens, the party retreated behind the populist traditions of the past and declared the Sungu Sungu to be the real rural people’s militia”.203 As the Sungu Sungu were integrated into the village security structure, CCM had successfully coopted this “tremendous burst of popular energy”.204

VI Transition to multiparty competition (1992-present)  

In the early 1990s, Tanzania embarked upon a process of political liberalisation. The latter involved far-reaching changes in political organisation including (1) the separation of CCM from the government; (2) the introduction of a multiparty system; (3) the de-linking of the trade unions and cooperatives from the ruling party; (4) and the granting of both freedom of press and freedom of association.205 Political liberalisation has also raised fundamental issues about the relation of the security forces to the civilian state.
Expectedly, the dismantling of the deeply entrenched one-party state was not free of tensions. On the Tanzania mainland, the unregistered Democratic Party has launched political propaganda and even attacks on Asian traders who often still “dominate profitable economic activities”.206 But the by far the most serious source of instability and violent conflict arises from the always imperfect union between the mainland and Zanzibar where politics has been increasingly conducted through an idiom of religion. Yet on the whole and once again, the country managed to maintain a remarkable degree of political stability. This enduring stability can be explained by CCM’s ability to retain its overwhelming dominance of the political system. While there is some evidence for reviving ethnic and regional loyalties and changing patterns of rent deployment, the ruling party still exerts an “immense gravitational pull”207 and maintains centralised control over patronage.
                                                 200 Zirker 1992, 114 201 Zirker 1992, 117 202 Campbell 1992, 101f. Cf. also Heald 2006. 203 Campbell 1992, 101f. 204 Campbell (1992, 102) claims that CCM’s success became evident “in the way the party called on this traditional army to form the guard of honour for leaders of the state at the Peasants Day in Shinyanga in July 1987”. Chris Peter (interview), by contrast, argues that CCM has tried to institutionalise the Sungu Sungu but has so far been unable to effectively get rid of this rivalling security system. 205 Tripp 2000, 197 206 Klugman et al. 1999, 96. 207 Kelsall 2003, 62

POLITICS
Political liberalisation was not driven by a strong popular movement but can better be described as “top-down democratisation”.208 According to Aili Tripp, democratisation from above was deemed inevitable for at least three reasons.209 First, the “democratizing winds sweeping Africa” generally made political reforms hard to avoid. By “jump-starting political liberalisation”, leaders sought to manage and control the reform process without endangering their own political power. Second, CCM did not want to compromise its relations with foreign donors who made continued support conditional on steps towards democratisation. Third, the ruling party hoped to be able to strengthen its links with the people that had been weakened throughout the 1980s (see above). A controlled process of political liberalisation was perceived as a suitable means to “enhance the CCM’s credibility” over the opposition.
In line with these initial calculations, CCM has indeed so far been able to assert its political hegemony in the era of multiparty politics. In the 1995 elections, the first multiparty elections in 35 years, CCM candidate Benjamin Mkapa was elected President winning 61.8% of the vote.210 CCM also won a large parliamentary majority with 186 seats equalling about 80% of the vote. The political opposition was divided into several parties, only four of whom managed to win parliamentary seats. The strongest opposition party was the Civic United Front (CUF) that won 24 seats, mainly because of its strength in Zanzibar where it won close to 50% of the vote. The second largest opposition party was the National Convention for Constitutional Reform (NCCR-Mageuzi), which won 16 seats. The Party for Democracy and Progress (CHADEMA) and the United Democratic Party (UDP) each won three opposition seats. CCM’s considerable margin of victory was even increased in subsequent elections, not least since many of those who initially left the ruling party to set up opposition parties have returned in the meantime. In 2000, Mkapa was comfortably reelected President, this time winning 71.7% of the vote. In Parliament, CCM won a total of 206 seats, while the 25 opposition seats were divided between the CUF (16 seats), the CHADEMA (4 seats), the Tanzania Labour Party (TLP) (3 seats) and the UDP (2 seats).211 In 2005, finally, the new CCM candidate and current President Jakaya Kikwete was elected with 80.3% of the vote. The elections for the Bunge (Parliament) again resulted in a one-sided victory for CCM that now secured 206 out of the 232 parliamentary seats. The opposition seats were divided between the CUF (19 seats), the CHADEMA (5 seats) as well as the TLP and UDP (1 seat each). In sum, “CCM’s stranglehold on the political system has not been significantly weakened by multi-party politics” whereby Tanzania’s one-party state has been transformed into a “dominant party system”.212
CCM’s overwhelming electoral dominance may seem puzzling, given the substantial loss of credibility that the ruling party had suffered throughout the 1980s. Yet, the ruling party’s enduring hegemony can be explained by a number of different factors.213 To begin with, as associational life in Tanzania has been crippled since independence and continues to be controlled by CCM in the post-1992 period (see below), the opposition parties have clearly suffered from the “absence of ready-made social bases” on which they could have built. Moreover, opposition parties have been unable to build a national presence, not least due to                                                  208 Hyden 1999, 154. This democratisation from above challenges the pattern identified by Bratton & van de Walle (1997) who argue that democratic transitions in Africa tend to originate in society rather than among elites (cited after ibid.). 209 Tripp 2000, 197f. 210 The following account of recent electoral developments is based on EIU 2007, 6ff. 211 Note that the NCCR-Mageuzi lost all of its 16 seats in the 2000 elections. 212 Kelsall 2003, 58. 213 Kelsall 2003, 59ff.

the country’s highly fragmented ethnic structure. In this context, CCM has found it relatively easy “stigmatise fledgling opposition parties with strong local bases as tribalist, thereby impeding their expansion into other areas”.214 Finally, and arguably most importantly, the CCM is not competing in elections on a level playing field as it still enjoys massive advantages in terms of party funding and controls access to key campaigning resources (e.g. stadiums). In the end, CCM’s enduring dominance has largely been driven by the party’s entrenchment within the state and its tight command over patronage. 215 In the words of one Tanzanian scholar, “(e)ventually one needs to go to the government to get a contract”.216
While multi-party democracy has not really taken root, there is also evidence that CCM has equally managed to maintain a tight grip on other elements of the political liberalisation process. According to Tripp, the government does not display any serious commitment “to adopt any measures beyond continuing to hold multiparty elections”. 217 Whereas the executive continues to dominate the judiciary and legislature thereby constraining political accountability, the ruling party and the state bureaucracy remain de facto closely intertwined. Furthermore, CCM has so far mostly been able to control the growing associational life in the country.218 While the number of registered NGOs has increased exponentially from 224 in 1993 to 8.360 in mid-1997 (women’s organisations having particular weight), the organisations often lack autonomy from the government and the dominant party. The mass organisations that had previously been tied to CCM experienced particular difficulties in emancipating themselves. The Organisation of Tanzanian Trade Unions (OTTU), for instance, was formed in 1991 to replace the party-controlled trade union but still remained under the tight control of the CCM-led government. Moreover, its leadership has been able to “insinuate itself” into the competing Tanzania Federation of Free Trade Unions (TFTU), which therefore also “remains only semiautonomous from the CCM”. 219 The case of the cooperative movement is only different at first sight. Even though it was de-linked from CCM through the Cooperatives Act of 1991, it is weak when compared with the pre-independence momentum and remains subject to increasing government re-regulation (see below). Many other of the “new generation organizations” are officially independent of the party and government, but also face “enormous pressures for cooptation”, e.g. by means of deregistration or stateinitiated umbrella organisations.
Altogether, CCM has clearly managed to maintain a high degree of centralised authority even in times of “political liberalisation”. The ruling party “is more a state party rather than a political party” – a “ruling machine designed to maintain a grip on power (…)”.220 At the same time, it has largely preserved its inclusive character and remains “broadly reflective of the country at large and manifests many of the same cleavages. The leadership is divided in its religious, regional, generational and ethnic identifications, and in its ideological persuasions”.221
                                                 214 According to Kelsall (2003, 60), this “happened to NCCR, which the CCM identified as a Chagga party, and the UDP, which was accused of being a vehicle for Sukuma interests”. 215 Superior access to state resources has made it easy for CCM to undermine opposition parties. As one informant argued, the “CCM is used to big victories. During the first five years [of the multiparty period] the government and CCM spent effort and resources to undermine small parties. They used the government to create disharmony in small parties” (Chris Peter, interview). 216 Muya, interview. Cf. also Hyden 1999, 144. 217 Tripp 2000, 198. 218 Tipp 2000, 199ff. 219 For details cf. Tripp 2000, 202 220 Baregu 2002, 71. 221 Kelsall 2003: 61 f.

Despite the maintenance of a centralised and inclusive ruling coalition, some observers have expressed concern that the introduction of multiparty politics “has revived ethnic and regional loyalties and heightened tensions around class, religious and other factional differences (…)”.222 And there are indeed some (albeit still minority) instances that point into this direction. Even though the constitutional amendment that legalised the multiparty system in 1992 prohibits political parties from tribal, religious and racial biases in order to protect national unity, several opposition parties have de facto organised along ethnic, regional or racial lines. As mentioned above, the NCCR-Mageuzi had their stronghold in the Kilimanjaro region, while the bulk of UDF’s followers were Sukuma. Also, while in the past TANU and later the CCM often ran its leading candidates in areas different from the regions of their ethnic origins,223 there is now a distinct trend for candidates to stand in their local areas. Religion has also become the basis of an appeal for votes,224 while one campaigner (Mtikila) beat up masses with the overtly racist slogan “Throw all Indians into the Oceans”.225 Especially worrying is a new trend that puts into question the strong notion of political citizenship that Nyerere had earlier fought for. Kaya argued that “currently, there are people arguing for uzawa (black indigenisation) in the economy.”226 CCM officials in the state have at times used citizenship to exclude potential potent political opponents.227 Last but not least, CCM is affected by the emergence of “regional blocs” among MPs.228 The “southern bloc”, based in the Southern Highlands, claims that northerners continue to hold key positions of power. The increased visibility of the “southern bloc” has in turn consolidated the emergence of other “blocs”, e.g. a “northern group” and a group of MPs from the Lake Zone.
Even though these instances remain minority and are unlikely to endanger political stability in the short- and medium-term, they nonetheless indicate a certain change in the way politics is conducted. As political discourse and campaigning has increasingly shown signs of appealing to the electorate on the grounds of ethnicity, regionalism, religion and racialism in marked contrast to the past, this raises questions about the long-term stability of the Tanzanian state.
ECONOMY
The extreme downturn of the 1980s was followed by macroeconomic recovery and stabilisation since the mid-1990s.229 Economic growth averaged 4.2% between 1996 and 2000, then picked up to 6.2% in 2001 and 7.2% in 2002, before moderating to 5.7% in 2003, and
                                                 222 Mmuya 2000, 71. Erdmann (2000, 93), by contrast, claims that the “chain of causation is exactly the reverse. It is precisely because the authoritarian regime could not provide for social and economic development, and because social and political integration was threatened, there remained only the attempt at democracy. The growing social tensions were only covered over by authoritarianism through violence and a clientelist distributive structure. Without the democratisation, sooner or later these tensions were going to erupt more violently and in ways that we have seen in other countries”. 223 “Julius Nyerere, Amir Jamal, Omar Muhaji and Oscar Kambona who stood, and won, in constituencies outside their home areas. Derek  Bryceson is a classic example of a white, Cambridge-educated big farmer standing in a working class area, the numerically largest constituency in the whole country, and still winning massively” (Kaya 2004, 165). 224 Kaya 2004, 164f. 225 Source? 226 Kaya 2004, 164 227 Oscar Kambona, who was Minister of Industry and Secretary General of TANU went into exile in 1967 in England after he fell out with Nyerere. When he wanted to return in 1992 to participate in forthcoming elections he was told he was not a citizen. Declaring people ‘non-citizens’ has also been used to gain access to business interests. During the debate on “indigenisation” after independence, Nyerere called it “camouflaged apartheid” (Chris Peter, interview). 228 Kelsall 2002, 606f. 229 The following account of economic performance is based on EIU 2007, 28ff.

then recovering to 6.7% in 2004 and 6.8% in 2005. Growth in real GDP per head (at 2000 prices) has been slower, rising from TSh199531 in 1994 to TSh265965 in 2004. The performance of the agricultural sector, on which the economy remains heavily dependent with a 45% share in GDP, remains mixed despite an output growth by over 5% in 2004-05. The manufacturing sector has struggled to grow over the past decade with its contribution to GDP falling from 9% in 1991 to 6.8% in 2005. Growth in the minding sector, by contrast, has been extremely rapid, at 27.4% in 1998, 9.1% in 1999, and then averaging 15.3% per year between 2000 and 2005.230 Even though the sector’s overall contribution to GDP remains small with only 3.8% in 2006, minerals have now replaced agricultural products in terms of export value. While earnings from traditional commodity exports (coffee, cotton, sisal, tea, tobacco, cashew nuts and cloves) fell from $436 million in 1997 to an average of $243 million between 2000 and 2003, revenues from mineral exports rose from just $26.4 million in 1998 to an estimated $823.9 million in 2006, dominated by gold, which is now Tanzania’s single largest export commodity.
Economic stabilisation in the 1990s was accompanied by new rounds of structural adjustment. Tanzania’s commitment to reform strengthened during Mkapa’s first term and then picked up considerably during his second term.231 Tanzania even became “a star reformer in the eyes of many donors” and prompted substantial inflows of foreign aid, which have accounted for around 40% of the government’s budget. Significantly, and in sharp contrast to the period between 1981 and 1992, adjustment now involved structural reform in public administration and the parastatal sector – the bastions of Tanzania’s elite bargain. Between 1992 and 1998, over 80.000 public servants were laid off (including thousands of ghost workers), whereby the total number of central government employees declined from 335.000 to 255.000.232 These initial retrenchments were deemed necessary mainly to maintain aid flows. The pace of public sector reform has clearly increased under Mkapa, especially since 2000.233 While Tanzania’s civil service is both smaller and better paid than previously, Mkapa has also cut the number of ministries in the cabinet.234 In the parastatal sector, structural reform has been even more pronounced. Out of the 413 public enterprises in the late 1980s (see above), as much as 200 had been privatised by 1998235 – a number that increased to 266 in 2003.236
The significance of these structural reforms should not be exaggerated since both public administration and the parastatal sector are still of considerable size and thereby continue to aliment existing patterns of elite accommodation. Nonetheless, Kelsall is certainly right to point out that the “state has shrunk and so is no longer sufficient to act as a vehicle for reproducing the elite in toto”.237 This structural change has been combined with the increasing liberalisation of the Tanzanian economy, which endangers the manifold commercial rents that had existed in the context of the highly statist economy. As a consequence, rent-seeking activities are changing (see below). On the one hand, elites are battling to regain control over imperfectly liberalised markets in order to protect rents. On the other, rent-seeking has at least partially started to shift away from the state and became increasingly entangled with private capital, sometimes in a highly localised context. These developments make the “elite bargain”
                                                 230 For details on the history and rising importance of Tanzania’s mineral sector cf. Chachage 1995. 231 EIU 2007, 29 232 Temu & Due 2000, 703f. 233 Kjaer 2004, 399ff. 234 Nonetheless, Kjaer (2004, 401) reports that even Mkapa “must distribute some patronage” through the creation of independent portfolios. 235 Temu & Due 2000, 694 236 Lockwood 2006, 111 237 Kelsall 2002, 610, original emphasis.

more difficult to manage and may – at least in the long term – have a centrifugal impact on the cohesion of the country’s ruling coalition.
The often imperfect liberalisation of markets under structural adjustment provides national and local elites with manifold opportunities for accumulation. The agricultural sector is a good case in point. In the early 1990s, the monopoly of state trading companies and cooperative unions for both export and food crops was broken, fertiliser subsidies were phased out and markets were opened to private traders.238 While the liberalisation of the markets of maize and other grains has been significant and sustained, export crop liberalisation remained highly imperfect, in particular for coffee, tobacco and cashew.239 In these cases, the initial liberal surge after 1993-94 was gradually undermined by (1) the reempowerment of export crop boards, which re-emerged as major actors in the marketplace;240 (2) the arbitrary treatment of farmers by local government, most notably through taxation; (3) the proliferation of sector policies that privilege the state as “initiator rather than facilitator, especially through foreign aid funded projects; and (4) continued government-guaranteed bank lending to select cooperative unions, which facilitated their partial recovery. The rationale underlying this anti-liberal behaviour of the political class and its clients in cooperatives and crop boards is clearly the preservation and perpetuation of lucrative agricultural rents.
The example of the coffee sector is illuminating in this respect.241 When the domestic coffee trade was opened to private traders and processors in 1994-95, foreign companies rapidly management to capture the Tanzanian coffee market, both domestically (trade and processing) and export.242 While the market share of cooperative unions fell from 83% in 1994-95 to just 26% in 1999-2000 (with only a few large unions being able to survive), independent local traders also lost ground – a disempowerment of Tanzanian interests that was accompanied by widespread popular discontent. Ponte shows that the government adopted a series of measures to “re-empower” local elite interests including the “manipulating of licensing rules, encouraging direct selling of coffee at auction by independent co-operatives and farmer groups, auction haggling, threats of tightening regulation and, above all, the period or remonopolisation of domestic coffee marketing in Kagera”.243 In the latter case, the ruling party not only rescued cooperatives from bankruptcy (by repaying its large debts and even backing new loans) but also banned private traders from operating in local markets (most of whom were agents of foreign companies) – a move that helped it to win the 2000 elections in the region. Developments in the Kagera region had repercussions at the national level in that the 2002-03 Coffee Act prohibits exporters from holding a domestic trading license. While Ponte convincingly argues that the Act was meant to reverse the “pattern of foreign ownership” (not least since private traders were re-allowed to buy domestic coffee), the re-regulation of                                                  238 Cooksey 2003, 69ff. After the Cooperative Societies Act (1991), only economically viable cooperatives would be registered, membership became voluntary and unions were given more autonomy and began setting their own prices. As the government stopped guaranteeing loans, banks started to be more cautious in lending (Ponte 2004, 626). 239 Cooksey (2003, 76) hypothesises that these differences can best be explained by the fact that “the price of maize – the food staple of the urban poor – is too important politically to be left to the mercies of marketing board and co-operative lobbyists. This political imperative does not apply to export crops, where national-local patronage policies and systematic rent-seeking can be more easily accommodated”. 240 Through the state marketing boards, the Minister of Agriculture retains draconian powers over crop production, marketing and taxation (cf. Cooksey 2003, 77ff.). 241 Another interesting example would be the cashew sector. Chachage & Nyoni (2001, cited after Kelsall 2003, 68) have shown how the cashew board, local government, traders and investors have manipulated markets to extract rents from cashew farmers. 242 Ponte 2004, 618ff. 243 Ponte 2004, 631

agricultural trade can equally be interpreted as a struggle by Tanzanian elites to regain control over rent deployment.
Another key site of struggles over rents are land issues, which have gained increasing salience in the recent period. A process of land accumulation by powerful elites began already during the villagisation period and gained momentum when the National Agricultural Policy (1982) introduced elements of private ownership of land (see above). According to Khan & Gray, the “new land laws that were enacted in 1999 were intended to address the competing pressures generated by the emergence of land markets on one hand and security of tenure for smallscale peasant farmers on the other (…).  Far from achieving these ends, the actual implementation of ‘land reform’ is not occurring under the clear guidance of the state but rather through the actions of private individuals engaging in illegal non-market transfers”.244 As ultimate land titles remain vested in the state, the government has been able to provide individuals – mostly foreign investors or Tanzanians of Arab and Asian origin – with large tracts of land.245 This form of patronage distribution has increasingly given rise to disputes with local farmers over water and land. Moreover, the recent wave of privatisations has opened new opportunities for rent deployment since influential politicians or bureaucrats may use their position to receive kickbacks or become joint partners in the privatised companies. The privatisation of parastatals has prompted increasing struggles and suspicion among elites, not least since half of the enterprises have landed in foreign hands.246.
A final, and particularly interesting example of newly emerging rent-seeking activities are the so-called District Development Trust Funds.247 Trust funds are non-state institutions that are run by local elites – sometimes retrenched civil servants – and seek to fund local development projects, most notably education. They finance themselves by taxing local populations and securing donations from patrons. Typically of ethnic nature, they constitute an important link between local and national elites. National elites patronise trust funds both to have access to local resources (e.g. minerals, cash crops, fisheries, tourism) and to ensure local political support, which has become increasingly important in times of multi-party politics. Local elites use the trust to finance education for their children, built local prestige or even misappropriate their revenues. Altogether, “local and national elites are linked, more closely than ever before, by a skein of reciprocities, the ideology to which is localism”.248
SECURITY
The introduction of multi-party politics brought about important changes in Tanzania’s civilian-military relationship. While 30 years of civilian-military integration had created a situation where it was becoming to a certain extent difficult to draw a line between CCM and the armed forces, the civilian and military spheres now had to be formally de-linked.249 Party structures in the army were abolished and army personnel are now prohibited from being involved in politics, that is, they can no longer join political parties.
One of the greatest worries about the political stability implications of the multiparty era is therefore the fate of the Tanzanian armed forces. But even though the end of the one-party
                                                 244 Khan & Gray 2006, 58f.  245 Kelsall 2003, 67 246 Muya, interview [check data] 247 Kelsall 2002, 611; 2003, 71. For a more detailed discussion cf. Kiondo 1995.  248 Kelsall 2002, 611 249 Omari 2002, 102f.

state seems to endanger Tanzania’s achievements in subordinating the army under civilian rule, there is strong evidence that so far “not much as changed”.250 Omari reports that the habit of co-opting army members into the civilian sphere has continued even after 1992. By 2002, 45% of regional commissioners and 20% of the district commissioners were still of military background. Also, many trop ranking military officers contested the 1995 elections under the umbrella of the CCM (even though required to resign their army duties with no possibility of returning afterwards. Altogether, the available evidence suggests that the majority of the TPDF still align with the ruling party and that civilian-military relations remain fairly healthy.251
Nonetheless, there is also evidence that armed forces are slowly becoming “a law unto itself (…). Recently, questions have been asked about the army’s involvement in gold mining, links between the Tanzanian military and Robert Mugabe’s regime in Zimbabwe, evidence of gun running in the Democratic Republic of Congo, not to mention the purchase of an extremely expensive military radar system. A dissident MP gave as his opinion on what the first of these issues suggested about state-military relations: No man wants his wife to go into business alone. If a man allows his wife to go into business, that suggests to me that he is no longer able to feed her”. 252 Furthermore, The central question remains whether the military organisations of the state have become institutionalised to the point that they can serve any political masters who occupy state offices. In case another political party gains power – no matter how unlikely this may seem at the moment – it will either have to purge all known CCM members in the armed forces or prepare to face hostilities.
Finally, there remain security worries related to the lack of capacity of the state’s police forces to ensure the safety of people in their everyday lives. The rise, and tolerance by the state, of Sungu Sungu organisations to enforce local law and order (see above) is an indication of the precariousness of security in the rural areas. At least to date, however, the violent conflicts that have occurred are related more to criminal activity or local disputes over land or grazing rights rather than taking on a more general political character.253 Even in urban areas, there has been a rapid expansion of private security forces. Tanzania still has only one police station for every 30-40,000 people. While these trends are worrying, by all accounts the police, while lacking in capacity, remain a centrally organised force with a significant degree of discipline.

VII Conclusion  

Summary of argument   to follow
We will consider our conclusions in light of Khan and Gray:
Khan & Gray 2006, 50f.: “While Tanzania has clearly achieved political stability since independence, this has not led overall to a higher level of state effectiveness as there is an important distinction between inclusive politics and strong party politics. The CCM party in Tanzania has historically been an inclusive organization that has been able to contain factional competition through its well-developed and inclusive institutional structures in a political                                                  250 Omari 2002, 102f. 251 Ibid.; Mukandala, interview. 252 Kelsall 2003: 62 253 Source?

context where political entrepreneurs are not very powerful given the low level of economic differentiation and development. Further, the low level of social differentiation and weak economic transition has paradoxically reduced the costs of political competition given the existence of an inclusive one-party state that is still effectively a one-party state despite the beginnings of multi-party competition. So far, at least, the ability of different factions in the business community and within the state to mobilize broader social groups to demand or protect rents has been very limited. Clientelist political corruption in Tanzania therefore ensures that the centre remains stable through internal redistribution to diffuse power networks even though the centre itself lacks the political power to act as a strong state”.      

VIII Bibliography  

Allen, Chris. 1995. ‘Understanding African Politics’, Review of the African Political Economy 65: 301-320. Baregu, Mwesiga. 1994. ‘The rise and fall of the one party state in Tanzania", in Jennifer A. Widner, J.A. (ed.), Economic Change and Political Liberalisation in Sub-Saharan Africa. Baltimore: The Johns Hopkins University Press, Baltimore, 158-81. Baregu, Mwesiga. 2002. ‘The State of Political Parties in Tanzania, 1999’, Paper presented at the International IDEA Conference on Towards Sustainable Democratic Institutions in Southern Africa, Gaborone, Botswana, 8–10 May. Bratton, Michael and van de Walle, Nicholas. 1997. Democratic Experiments in Africa: Regime Transitions in Comparative Perspective. Cambridge: Cambridge University Press. Campbell, Horace. 1992. ‘The politics of demobilisation in Tanzania: Beyond nationalism’, in Horace Campbell and Howard Stein (eds.), Tanzania and the IMF. The dynamics of liberalisation. Boulder: Westview Press, 85-108. Chachage, Seithy L. 1995. ‘The meek shall inherit the earth but not the mining rights: The mining industry and accumulation in Tanzania’, in Peter Gibbon (ed.), Liberalised development in Tanzania. Uppsala: Nordiska Afrikainstitutet, 37-108. Chachage, Seithy L. and Nyoni, Joyce. 2001. ‘Economic Liberalisation and the Cashewnut Industry in Tanzania’, Paper presented at the Open Forum on the Cashewnut Industry in Tanzania, Dar es Salaam.  Collier, Paul and Hoeffler, Anke. 2002. ‘On the Incidence of Civil War in Africa’, Journal of Conflict Resolution 46(1): 13-28. Cooksey, Brian. 2003. ‘Marketing reform? The rise and fall of agricultural liberalisation in Tanzania’, Development Policy Review 21(1): 67-91. Costello, Matthew J. 1996. ‘Administration triumphs over politics: The transformation of the Tanzanian state’, African Studies Review 39(1): 123-148.

Coulson, Andrew. 1982. Tanzania: A political economy. Oxford: Oxford University Press. Dashwood, Hevina S. and Pratt, Cranford. 1999. ‘Leadership, participation, and conflict management: Zimbabwe and Tanzania’, in Taisier M. Ali and Robert O. Matthews (eds.), Civil wars in Africa: Roots and resolution. Montreal: McGill-Queen's University Press, 223256. DiJohn, Jonathan. 2008. ‘Conceptualising the Causes and Consequences of Failed States: A Critical Review of the Literature’, Crisis States Working Papers Series 2 25, London: Crisis States Research Centre, London School of Economics. EIU (Economist Intelligence Unit). 2007. Tanzania. Country profile 2007. London: EIU. Elbadawi, Ibrahim and Sambanis, Nicholas. 2000. ‘Why are there so many civil wars in Africa? Understanding and preventing violent conflict’, Journal of African Economies 9(3): 244-269. Ellis, Frank. 1983. ‘Agricultural marketing and peasant-state transfers in Tanzania’, Journal of Peasant Studies 10(4): 214-242. Erdmann, Gero. 2000. ‘A cautious promise for stable democratisation?’, in Ulf Engel, Gero Erdmann and Andreas Mehler (eds.), Tanzania revisited: Political stability, aid dependency and development constraints. Hamburg: Institute of African Affairs, 93-96. Harnevik, Kjell J., Kjaerby, Finn, Meena, Ruth, Skarstein, Rune and Vuorela, Ulla. 1988. Tanzania. Country Study and Norwegian aid review. Bergen: Centre for Development Studies, University of Bergen. Harnevik, Kjell J. 1993. Tanzania. The limits to development from above. Uppsala: Nordiska Afrikainstitutet.  Suzette Heald. 2002. ‘Domesticating Leviathan: Sungusungu groups in Tanzania’, Crisis States Working Papers Series 1 16. London: Crisis States Research Centre, London School of Economics. Heald, Suzette. 2006. ‘State, law, and vigilantism in Northern Tanzania’, African Affairs 105(419): 265-283. Hofmeier, Rolf. 1993. ‘Tanzania’, in Dieter Nohlen and Franz Nuschler (eds.), Handbuch der Dritten Welt. Bonn: Dietz. Hofmeier, Rolf. 1997. ‘Tanzania: "Friedensoase" in der subsaharischen Krisenzone‘, in Volker Matthies (ed.), Der Gelungene Frieden: Beispiele und Bedingungen erfolgreicher friedlicher Konfliktbearbeitung. Bonn: Stiftung für Entwicklung und Frieden, 151-174. Hyden, Goran. 1980. Beyond Ujamaa in Tanzania. Underdevelopment and an uncaptured peasantry. London: Heinemann. Hyden, Goran. 1999. ‘Top-down democratisation in Tanzania’, Journal of Democracy 10(4): 142-155. Iliffe, John. 1979. A modern history of Tanganyika. Cambridge: Cambridge University Press. Kaya, Hassan O. 2004. ‘Electoral Violence, Political Stability and the Union in Tanzania’, African Journal on Conflict Resolution 4(2): 145-69.

Kelsall, Tim. 2002. ‘Shop windows and smoke-filled rooms: governance and the repoliticisation of Tanzania’, Journal of Modern African Studies 40(4): 597-619. Kelsall, Tim. 2003. ‘Governance, democracy and recent political struggles in mainland Tanzania’, Commonwealth & Comparative Politics 41(2): 55-82. Khan, Mushtaq and Gray, Hazel. 2006. ‘State weakness in developing countries and strategies of institutional reform – operational implications for anti-corruption policy and a case-study of Tanzania’. Kjaer, Anne M. 2004. ‘Old brooms can sweep too! An overview of rulers and public sector reforms in Uganda, Tanzania and Kenya’, Journal of Modern African Studies 42(4): 389-413. Kiondo, Andrew. 1992. ‘The nature of economic reforms in Tanzania’, in Horace Campbell and Howard Stein (eds.), Tanzania and the IMF. The dynamics of liberalisation. Boulder: Westview Press, 21-42. Kiondo, Andrew. 1995. ‘When the state withdraws: Local development, politics and liberalisation in Tanzania’, in Peter Gibbon (ed.), Liberalised development in Tanzania. Uppsala: Nordiska Afrikainstitutet, 109-176. Klugman, Jeni, Neyapti, Bilin, and Stewart, Frances. 1999. Conflict and growth in Africa. Vol. 2: Kenya, Tanzania and Uganda. Paris: OECD. Lindemann, Stefan. 2008. ‘Do inclusive elite bargains matter? A research framework for understanding the causes of civil war in Sub-Saharan Africa’, Crisis States Discussion Papers Series 15, London: Crisis States Research Centre, London School of Economics. Lockwood, Matthew. 2006. The state they’re in. An agenda for international action on poverty in Africa. London: ITDG Publishing. Lofchie, Michael. 1993. ‘Trading Places: Economic Policy in Kenya and Tanzania‘, in Thomas M. Callaghy and John Ravenhill (eds.), Hemmed In: Responses to Africa’s Economic Decline. New York: Columbia University Press, 398-462. Lupogo, Herman. 2001. ‘Civil-military relations and political stability’, African Security Review 10(1). Maliyamkono, Ted L. and Bagachwa, Mboya S. 1990. The Second Economy in Tanzania. London: James Currey. Martin, Denis-Constant. 1988. Tanzanie. L’invention d’une culture politique. Paris: Presses de la Fondation Nationale des Sciences Politiques. McGowan, Patrick J., and Wacirah, H. K. 1974. ‘The evolution of Tanzanian political leadership’, African Studies Review 17(1): 179-204. Mmuya, Max. 2000. ‘Democratisation, party politics and elections in Tanzania’, in Ulf Engel, Gero Erdmann and Andreas Mehler (eds.), Tanzania revisited: Political stability, aid dependency and development constraints. Hamburg: Institute of African Affairs, 71-92. Mpangala, Gaudens P. 1999. Peace, conflicts, and democratisation process in the Great Lakes Region: The experience of Tanzania. Dar es Salaam: Institute of Development Studies, University of Dar es Salaam. Nyerere, Julius. 1968. ‘Socialism and rural development’, in Julius Nyerere, Ujamaa-Essays on socialism. Oxford: Oxford University Press. Omari, Cuthbert K. 1995. ‘The management of tribal and religious diversity’, in Colin Legum and Geoffrey Mmari (eds.), Mwalimu: influence of Nyerere. London: James Currey, 23-31.

Omari, Abidallah H. 2002. ‘Civil-military relations in Tanzania’, in Rocky Williams, Gavin Cawthra and Diane Abrahams (eds.), Ourselves to Know. Civil-Military Relations and Defence Transformation in Southern Africa. Pretoria: Institute for Strategic Studies, 89-106. Polanyi, Karl. 1944. The Great Transformation: the political and economic origins of our time. Boston: Beacon Press. Ponte, Stefano. 2004. ‘The politics of ownership: Tanzanian coffee policy in the age of liberal reformism’, African Affairs 103(413): 615-633. Pratt, Cranford. 1976. The critical phase in Tanzania, 1945-1968: Nyerere and the emergence of a socialist strategy. New York: Cambridge University Press. Putzel, James. 2002. ‘Politics, the state and the impulse for social protection: the implications of Karl Polanyi's ideas for understanding development and crisis’, Crisis States Working Papers Series 1 18. London: Crisis States Research Centre, London School of Economics. Putzel, James. 2007. ‘A Proposal for Cross-Country Comparative Research: Development as State-Making Component of CSRC Agenda’, unpublished manuscript. Shivji, Issa G. 1976. Class Struggles in Tanzania. London: Heinemann. Shivji, Issa G. 1990/ “Notes on the Crisis o f Ideological Hegemony”, TAAMULI, vol.1, no.s 1&2. Stein, Howard. 1992. ‘Economic policy and the IMF in Tanzania: Conditionality, conflict, and convergence’, in Horace Campbell and Howard Stein (eds.), Tanzania and the IMF. The dynamics of liberalisation. Boulder: Westview Press, 59-84. Stevens, Mike. 1994. ‘Public expenditure and civil service reform in Tanzania’, in David L. Lindauer and Barbara Nunberg (eds.), Rehabilitating Government: Pay and Employment Reform in Africa. Washington, DC: World Bank, 62-81. Stewart, Frances. 2000. ‘Crisis Prevention: Tackling Horizontal Inequalities’, Oxford Development Studies 28(3): 245-262. Swai, Col F. 1991. ‘The politicisation of the Tanzania Defence Forces’, in Jeannette Hartman (ed.), Re-Thinking the Arusha Declaration. Copenhagen: Centre for Development Research, 94-104. Temu, Andrew E., and Due, Jean M. 2000. ‘The business environment in Tanzania after socialism: challenges of reforming banks, parastatals, taxation and the civil service’, The Journal of Modern African Studies (38)4: 683-712. Tripp, Aili M. 2000. ‘Political reform in Tanzania: The struggle for associational autonomy’, Comparative Politics 32(2): 191-214. Tripp, Aili M. 1997. Changing the Rules: The Politics of Liberalization and the Urban Informal Economy in Tanzania. Berkeley: University of California Press.  Van de Walle, Nicolas. 2001. African Economies and the Politics of Permanent Crisis, 19791999. Cambridge: Cambridge University Press. Uvin, Peter. 1998. Aiding Violence: The Development Enterprise in Rwanda. West Hartford: Kumarian Press.


Zirker, Daniel. 1992. ‘The preservation of civilian rule in Tanzania’, in Constantine P. Danopoulos (ed.), Civilian rule in the developing world. Democracy on the march? Boulder: Westview Press, 107-123.